HomeMy WebLinkAboutMINUTES - 06031997 - D7 D.1
TO: BOARD OF SUPERVISORS
FROM: William Walker, MD ' ° 's
Health Services Director
Sra'ao��.
DATE: May 21, 1997
SUBJECT: Tosco Avon Refinery Incident of January 21, 1997
SPECIFIC REQUEST(S)OR RECOMMENDATION(S)&BACKGROUND AND JUSTIFICATION
Recommendation:
ACCEPT reports from the Health Services Department, Hazardous Materials Division Deputy
Director and a Tosco Avon Refinery Representative on the findings of the incident investigation
of the January 21, 1997 hydrocracker incident.
Background:
At approximately 7:42 p.m.on Tuesday,January 21, 1997,an explosion and fire in the hydrocracker
unit at Tosco's Avon Refinery caused the fatality of one Tosco employee and injured 46 others.
Incident investigations including root causes analyses were conducted by both Contra Costa County
Health Services Department, Environmental Division staff and Tosco Avon Refinery staff. The
contributing and root causes of this incident will be summarized.
Fiscal Impact•
None.
CONTINUED ON ATTACHMENT: YES SIGNATURE
--r
RECOMMENDATION OF COUNTY ADMINISTRATOR RECOMMENDATION OF BOARD COMMITTEE
APPROVE OTHER
SIGNATURE(S):
ACTION OF BOARD ON June 3, 1997 APPROVED AS RECOMMENDED X OTHER X
Please see Addendum (Attached) for a list of speakers and additional Board action.
VOTE OF SUPERVISORS
I HEREBY CERTIFY THAT THIS IS A TRUE
X UNANIMOUS (ABSENT ---------- ) AND CORRECT COPY OF AN ACTION TAKEN
AND ENTERED ON THE MINUTES OF THE BOARD
AYES: NOES: OF SUPERVISORS ON THE DATE SHOWN.
ABSENT: ABSTAIN:
Contact Person:Laura Brown 646-2286
§upervisors ATTESTED June 3, 1997
CC: County Administrator PHIL BATC 0 , LERK OF THE BO RD OF
Health Services Director SUPE V D COUNTY ADT
_WVRA
Hazardous Materials Division(via HSD)
Hazardous Materials Commission (via HSD)
John E. Miller, Tosco Avon Refinery BY.
Jim Payne, OCAW Representative
Donald
Wamp er
Donald R. Brown, CBE Representative
D.7
ADDENDUM
Item D.7
June 3, 1997
Subsequent to the presentation of the report from the Health Services Department
(Attached) and the report from Tosco Avon Refinery (Attached), Chairman
DeSaulnier invited the public to comment on the issues and the following persons
spoke:
Donald R. Brown, Communities for a Better Environment (CBE), 1801
Sonoma Boulevard Suite 117, Vallejo;
Tom Lindemuth, Hazardous Materials Commission, 501 Daisy Place,
Pleasant Hill;
Jim Payne, Oil Chemical and Atomic Workers' Union (OCAW), P.O.
Box 428, Martinez;
Anne Bouguennec, 2825 Parkway, Martinez; and
Loren Freeman (No address provided on speaker form).
All persons desiring to speak having been heard, Chairman DeSaulnier ended public
comment on the matter.
The Board members further discussed the issues and took the following action:
1. APPROVED the recommendation, as set forth on the attached Board order, to
accept reports from the Health Services Department, Hazardous Materials
Division Deputy Director, and Tosco Avon Refinery on the findings of the
incident investigation of the January 21, 1997, hydrocracker incident;
2. REQUESTED that prior to the July start-up date of the hydrocracker unit,
Tosco Avon Refinery representatives report to the Board of Supervisors on the
implementation of the twelve recommended responses, as delineated in their
May 29, 1997, incident investigation report;
3. REQUESTED that representatives from the Oil, Chemical and Atomic
Workers' Union (OCAW) and Communities for a Better Environment (CBE)
confer with the Chair, Board of Supervisors, in order to coordinate a schedule
for submitting reports or comments to the Board regarding the hydrocracker
incident and investigation; and
4. REFERRED to the Hazardous Materials Commission for consideration the
matter of exploring the utilization of technologically advanced refinery
instrumentation for the early detection of hydrocracker "hot spots".
cc: Supervisors
County Administrator
Health Services Director
- Hazardous Materials Division (via HSD)
- Hazardous Materials Commission (via HSD)
John E. Miller, Tosco Avon Refinery
Jim Payne, OCAW Representative
Donald R. Brown, CBE Representative
JOHN E.MILLER
Vice President,Refining
General Manager Avon/Rodeo
' San Francisco Area Refinery(SFAR)
at Rodeo
1380 San Pablo Avenue
ir 94572-1299
TOSCO Direct Dial:(10)245-4415
Facsimile:(510)245-4586
May 29, 1997
TO: Contra Costa County Board of Supervisors
Safety at the Tosco Avon Refinery is our responsibility and one that the people who work here take very seriously.
We deeply regret the terrible tragedy which occurred on January 21,1997 and sincerely apologize to all those who
were hurt, affected, or concerned by the incident. No one wants to understand what happened, why it happened,
and how to prevent it from ever happening again more than we do.
The attached report is the result of over four months work by a team of very experienced men and women dedicated
to investigating the explosion and fire at the Avon Refinery, which cost our operator, Michael Glanzman, his life.
The investigation team was comprised of representatives-from many different refinery departments including an
OCAW Health and Safety Representative. The Tosco investigation team concluded that the cause of the explosion
and fire was a pipe rupture in the Hydrocracker Unit at the Avon Refinery. This pipe rupture was caused by a rapid
high temperature increase ("excursion") inside the unit, and by the fact that the unit was not depressured as
required by the emergency procedures. The investigation team has developed recommendations that address all
the causes of the high temperature excursion and lack of depressuring. We are implementing all of these
recommendations. Specifically, we are installing state-of-the-art-controls and instrumentation and an automatic
depressuring system, and we are re-training our operators on instrumentation and procedures, as well as other
actions detailed in the attached report.
The investigation team's mission was to determine the incident's causes and contributing factors, and to develop
recommendations for preventing anything like it from ever happening again. The team gathered all data, both
physical and technical, established the facts and analyzed all of the data. They had the full support and resources
of all of Tosco behind them, and were able to share information and receive assistance from many resources from
outside of the company. This included technical consultants, and specialists from the various regulatory agencies
who were on-site daily conducting a parallel investigation, including the Contra Costa County Health Services
Department, CAL OSHA (the lead agency due to a worker fatality), the Bay Area Air Quality Management District
and Federal EPA.
The Hydrocracker Report is comprehensive, very detailed, and by its nature, very technical. My staff and I will meet
with you to discuss the incident and report in detail. We are available to speak with you or your staff to answer your
questions over the telephone, to arrange a visit for you and a tour of the refinery or any combination of the above.
We are eager to share this information. The events which led to the explosion were unprecedented in the 34 year
history of the Avon Refinery Hydrocracker. None of our operators, all of whom are very experienced, had ever seen
anything like it before. Tosco operators and supervisors who worked with Michael Glanzman intend to
communicate our findings throughout the industry to help prevent a similar accident from ever happening anywhere.
Sincerely,
John E. Miller
Vice President and General Manager
San Francisco Area Refineries
JEM/c
(JEM-6997)
05i29i97 THt' 18:28 FAX 510 370 3397 COliMUNITI RELATION ltd 001
3° CONSIDER WON
TOSCO REFINING COMPANY
CLERK BOARD OF SUPERVISORS
" "RA COSTA CO.
FOR IMMEDIATE RELEASE CONTACT: Linda Saltzman
Tosco Refining Company
510/372-3093 (office)
510/228-1220 (24 hours)
TOSCO RELEASES REPORT ON JANUARY 21, 1997 ACCIDENT INVESTIGATION
. .t
MARTINEZ, CA (Thursday, May 29, 1997) -Tosco Refining Company today issued a report
detailing the findings of its extensive investigation into the January 21 explosion and fire in the
hydrocracker unit at the company's Avon Refinery. Nydrocracker units convert heavier oils to gasoline
and diesel products.
Among the findings of the-report:
' • The explosion was traced to a pipe rupture caused by a rapid high temperature increase in a
section of the hydrocracker. All possible causes have been identified and analyzed, although
the cause of the increase could not be conclusively determined after the explosion. All
possible causes have been addressed.
• Tosco found the incident to be highly unprecedented in the 34-year history of the
hydrocracker unit, outside the realm of its experience in the refining industry, and that of the
o - qualified operators on duty the night of January 21.
• A contributing cause of the incident was the fact that the unit was not depressured in
response to the temperature increase, as specified in the hydrocracker's emergency
procedures for safely controlling rapid high temperature increases. Employees did not
depressure the unit for a variety of reasons, most notably*
— The incident happened quickly (7Y2 minutes) and was unlike anything they had
previously experienced.
The chain of events and temperature readings the operators observed were not
what they expected to see, based on prior experience.
— They questioned whether the temperature indications shown on certain control room
instruments were accurate, because of previous computer malfunctions with similar
instruments.
During the investigation, Tosco found that none of the instruments malfunctioned that ...
night and all of the data the operators saw prior to the explosion were correct.
In response to the rapid temperature increase, an operator performed a routine check of
equipment outside the control room.
• While the operator was performing his equipment check, a pipe ruptured, causing the
explosion and fire which caused the employee's death.
Z
• -MORE-
-1-
ppKrtEes.docl
Vaai 4V.,V inL! 10. rV FAA J1V 01 V 0001 1VJ4V \.VU11R L'�:�111 l�ii L-111 V1y .ZJ vvr
TOSCO REFINING COMPANY
TOSCO RELEASES REPORT ON JANUARY 21, 1997 ACCIDENT INVESTIGATION (continued)
The Tosco report provides a minute-by-minute recounting of events leading up to the explosion,
and action taken by operators and emergency response personnel at the first sign of trouble.
The 38-page report was delivered today to the Contra Costa County Board of Supervisors and
other state and federal agencies. Tosco Refining Company officials, lead by company president Dwight
Wiggins, will present detailed findings of the report to the Board at its regular meeting in Martinez this
Tuesday, June 3. At that time, a separate report, prepared by the Contra Costa County Health Services
Department (HSD) will also be presented and discussed.
HSb was one of many regulatory agencies which investigated the January 21 incident_ Other
agencies included the California Department of Occupational Health and Safety (Cal-OSHA), the Bay
Area Air Quality Management District, and the U. S. Environmental Protection Agency. Tosco provided
each of the investigating agencies with documents, tours, briefings, and temporary on-site facilities as
well as arranged for Cal-OSHA and HSD staff to interview employees.
"We assume complete responsibility for safety at the Avon Refinery," said John Miller, General
Manager of Tosco's Avon Refinery. "Through this detailed report, we hope to fully educate our
employees, others in the industry, regulatory agencies, and the public about the findings of our
investigation. We want to make sure that something like this will never happen again."
"This event has affected us very deeply: we lost one of our employees, others were injured, and
we have created concern in the community," said Miller. "We are committed to the safest possible
operation and to rebuilding the trust which we have developed through more than 80 years of operations
of the Avon Refinery in Contra Costa County.'
Tosco will implement a series of corrective actions that will prevent such an incident from
reoccurring. These include:
+ installing state-of-the-art controls and instrumentation
• installing an automatic depressuring system
• installing state-of-the-art reactor internals
» re-training operators on instrumentation and procedures
In addition, Miller noted that the company has already implemented a number of improvements
to its emergency response access and notification protocols.
"We look forward to discussing these corrective actions and improvements with the County and
the public and welcome their input," said Miller.
-END-
-2-
(HDKPRESS.000
CONTRA COSTA COUNTY HEALTH SERVICES
SUMMARY REPORT TO THE BOARD OF SUPERVISORS
ON THE INVESTIGATION OF THE CAUSES OF THE
TOSCO AVON REFINERY INCIDENT OF 1-21-97
William H.Alton
Laura L. Brown
Contra Costa County Health Services Department
May 29, 1997
EXECUTIVE SUMMARY
Between 7:41 and 7:42 p.m. on Tuesday,January 21, 1997, an explosion and fire in the
hydrocracker unit at Tosco's Avon Refinery caused the fatality of one Tosco employee
and injured 46 others.
Contra Costa County Health Services Department Environmental Division(HSD)
conducted an incident investigation that included employee interviews and the review of
reports and records.
The root causes, as determined by HSD, of this incident are a hot spot in bed 4 of Reactor
3 created a temperature runaway coupled with the lack of training to provide the
operations group with the understanding of the need for prompt reaction to any unusual or
conflicting temperature indication.
INCIDENT INVESTIGATION
Cal-OSHA conducted an inspection to determine site compliance with the Process Safety
Management(PSM) standard. This federal and state law establishes procedures for
process safety management that will protect employees by preventing or minimizing the
consequences of chemical accidents involving highly hazardous chemicals. Cal-OSHA
was the lead agency in the investigation and all information requests for Tosco were
routed through Cal-OSHA. A final report is expected in July 1997; if citations are issued,
they will be issued at that time.
The Cal-OSHA Bureau of Investigations is conducting a criminal investigation.
HSD, the Federal EPA(Region IX and Washington D.C. staff members), and the Bay
Area Air Quality Management District(BAAQMD) staff conducted incident
investigations to determine causes.
The investigation spanned many weeks and included a number of employee (operator and
management) interviews by Cal-OSHA representatives to determine the sequence of
events leading up to the incident. Cal-OSHA representatives supplied information from
the employee interviews to members of the other participating agencies. Thousands of
records were reviewed which included maintenance inspection reports, operator logs,
temperature logs,management of change documentation,training records, and process
hazards analyses.
Page 1
The HSD analysis is based on information below plus HSD interviews with Tosco
employees.
DESCRIPTION OF UNIT AND EVENT CHRONOLOGY
Between 7:41 and 7:42 p.m. on Tuesday,January 21, 1997, an explosion and fire
occurred in the hydrocracker unit at Tosco's Avon Refinery. The material involved in the
explosion and fire was the reaction product from the hydrocracker stage 2, number 3
reactor. This product consisted of butane, light gasoline, heavy gasoline, gas oil and
hydrogen. The product was released from a rupture in the reactor outlet piping.
Additionally, Tosco reported that 13 pounds of friable asbestos was released during the
incident from insulation on equipment in the area of the explosion.
Hydrocracker Unit Description
A hydrocracker converts refinery gas oils (diesel and heavier oils)to gasoline blending
stocks and diesel. Refinery gas oils are initially treated with hydrogen(hydrotreated) in
three first stage fixed bed catalytic reactors (stage 1). The product is then sent through a
series of separators and strippers to remove sulfur, nitrogen, and other impurities1. The
treated gas oils are mixed with hydrogen,heated and hydrocracked2 in three second stage
fixed bed catalytic reactors (stage 2). This produces a mixture of products which are
separated into diesel and gasoline in an adjacent gas processing unit.
Catalyst in the reactors causes cracking, which is a reaction that requires heat
(endothermic). As the cracking process takes place,the hydrogen saturates the
molecules, which in turn generates heat(exothermic). The overall reaction is exothermic,
which causes reaction temperatures to increase as the gas oil/hydrogen mixture flows
through the catalyst. Reactor inlet temperatures are controlled by the addition of
hydrogen that has been heated in a hydrogen trim furnace. Temperature rise in the
catalyst is controlled by dividing the catalyst into 5 separate beds and injecting cold
quench hydrogen between the beds.
Impurities such as nitrogen could de-activate second stage catalyst.
2 Hydrocracking is catalytic cracking of gas oil with high pressure hydrogen supplied from an associated
hydrogen manufacturing plant.
Page 2
Unit Instrumentation
Prior to December 1995, reactor bed temperatures were monitored and alarms were
generated and recorded on a data logger. The center point temperatures for the bed inlets
and outlets were also displayed on strip chart recorders. Reactor control for the hydrogen
quench in stage 2 was achieved with Moore flow controllers using the same bed inlet
temperature points as displayed on the strip charts. There was a total of 40
thermocouples in five beds and 1 thermocouple for the reactor inlet piping and 1 for the
reactor outlet piping for each of the three stage 2 reactors.
In December 1995, Tosco began a project to upgrade the temperature monitoring system
for the hydrocracker stage 2. The upgrade project was to provide an increased number of
temperature points monitored for the stage 2 reactors and provide temperature monitoring
on a Foxboro intelligent automation(I/A) computer monitoring and control system.
During a January 1996 turn-around, Tosco installed additional thermocouples that more
than doubled the number in each reactor for a total of 96 thermocouples per reactor.
Local temperature indication panels were installed under each of the stage 2 reactors.
This resulted in 40 thermocouples in five beds, 1 thermocouple for the reactor inlet
piping, 1 for the reactor outlet piping continuing to be displayed and monitored in the
control room on the data logger, and 56 different points being available at the local
temperature indication panels for each of the stage 2 reactors.
On January 10, 1997, 40 thermocouples for each of the stage 2 reactors were transferred
from the data logger to the I/A system for temperature indication. (Note: The 56 points
at the local temperature indication panels at each of the stage 2 reactors remained.)
A tube leak on a reactor heat exchanger in stage 1 of the unit caused an unplanned
shutdown for repair from January 12 through January 16, 1997. On the morning of
January 16, 1997,the hydrocracker was started up with stage 2 reactor temperature
monitoring on the I/A system.
According to interviews with the operators and Tosco management, operators complained
that the I/A system was malfunctioning on January 19 and January 20. After
investigating the problems, it was determined that the temperature readings on the I/A
system were malfunctioning. The decision was made by management to switch from
temperature indication on the I/A system back to the data logger and continue the unit
start-up. This resulted in the stage 2 temperature indication and control being the same as
Page 3
it was prior to January 10, 1997 (i.e., data logger). The Moore controller system was
unchanged and remained functional.
The unit was also equipped with an emergency depressuring system. The 100 pound per
minute depressure would be automatically activated on the loss of the recycle
compressor, or manually activated any time a partial depressuring of the unit might be
desired. Procedures require manual activation of the 300 pound per minute depressure in
the event of a temperature runaway or if any temperature point in the reactor exceeds
800° F. The 300 pound per minute depressure shuts-off the oil feed,make-up hydrogen,
the hydrogen trim furnace, and depressures the hydrocracker to the refinery flare system.
Chronology of Events Leading to the 1/21/97 Incident
Late in the day on January 20, 1997, a leak occurred at an exchanger flange in stage 1 of
the hydrocracker. Feed was pulled from the reactor so that the leak could be repaired.
The feed was shifted to the remaining two reactors in stage 1. The increase in feed
through two reactors caused a very high off-test nitrogen content in the feed to stage 2.
From the operators' logs it is known that the catalyst in the stage 2 reactors became "de-
activated"3 on the morning of January 21, 1997. Tosco management decided to continue
running with de-activated catalyst4. During the day shift on January 21, 1997 (6 a.m. to 2
p.m.),the temperature and pressure readings in the stage 2 reactors were normal.
Nitrogen readings for stage 2 feed were taken throughout the day. By 5:40 p.m. that
evening,the nitrogen readings dropped to within the target specification range for
nitrogen in the feed to stage 2.
The operating plan for the evening specified increasing temperatures at stage 1 to reduce
the nitrogen content,then increase feed rates when possible. In addition,the temperatures
in stage 2 were to be gradually increased to restore catalyst activity. From a review of the
strip charts,the operators were properly executing this plan.
s Absorbed nitrogen and other impurities on the surface causes de-activation of the catalyst thus reducing
the hydrocracking reaction rates.
4 De-activated catalyst can be re-activated by either pulling the oil feed to the unit and purging the catalyst
beds in each reactor with hot hydrogen("hot sweeping")or by increasing the feed temperatures until the
impurities are removed from the catalyst. It is a common practice in the industry to run feed with de-
activated catalyst and remove nitrogen slowly by the latter method.
Page 4
At approximately 7:34 p.m.,the stage 2 reactor 3 beds 4 and 5 (bottom beds of the
reactor)temperature data logger began displaying what the operators believed to be
incorrect and erratic readingss. Operators reported seeing digital temperature readings on
the data logger that bounced low, high, and back again; some readings exceeded the
established upper operating limits. A temperature alarm activated on the.data logger for a
bed 4.outlet temperature monitoring point. A high bed 5 inlet temperature also displayed
on the data logger. The temperature data as displayed on the corresponding control room
strip chart for bed 4 showed a normal temperature due to the fact that the point that
alarmed on the data logger is not the control point for the corresponding strip chart
recorder. However,the bed 5 inlet temperature increase was displayed on the control
room strip chart. At about the same time, a sudden temperature rise of over 150 degrees
at the inlet of reactor 3 bed 5 caused the hydrogen quench valve to open fully.
By approximately 7:35 p.m.,the bed 4 outlet and bed 5 inlet temperatures (as shown on
the data logger) started dropping in response to the quench. At 7:36 p.m., the reactor
outlet temperature increased 9 degrees.
Between 7:36 and 7:37 p.m., the hydrogen quench to bed 5 was reduced in order to
maintain temperature in the reactor. At this same time,the Operators were concerned
about the trim furnace firing. A series of manual moves were made to the hydrogen trim
furnace.
Sometime between 7:34 and 7:38 p.m., the outside operator on duty that evening left the
control room and went to the bottom of reactor 3 in stage 2 to read the temperatures at the
local panel for comparison to the control room readings. The outside operator's response
on the local temperature readings came over the radio garbled.
At approximately 7:37 p.m.,while stage 2 reactor 3 of the hydrocracker was experiencing
what the operators believed was a temperature monitoring problem,the hydrogen plant
operator informed the hydrocracker stage 1 and stage 2 operators that the hydrogen make-
up feed to the hydrocracker had dropped to zero. Excess hydrogen was being re-directed
to the header/flare system to prevent overpressuring of the hydrogen plant.
5 For bed 4 there are five digital bed outlet and five digital bed inlet temperature indication points on the
temperature logger. For bed 5,there are four digital bed outlet and one digital bed inlet temperature
indication points on the temperature logger. One bed inlet point for each bed serves as the source of the
strip chart temperature recorder and as the control point for the associated hydrogen quench.
6 Make-up hydrogen replaces hydrogen that is consumed during the hydrocracking reaction.
Page 5
Between 7:38 p.m. and 7:40 p.m. the reactor 3 bed 5 strip chart and data logger outlet
temperatures and the reactor inlet strip chart and data logger temperatures showed a
sudden increase and some values defaulted to zero due to being high off-scale.
Between 7:41 and 7:42 p.m., an explosion occurred and the operators immediately
activated the 300 pound per minute depressuring system and began responding to the
incident. The explosion resulted in the death of the outside operator.
Once the fire was extinguished, it was found that the discharge line on the reactor 3 outlet
leading into a heat exchanger had ruptured releasing flammable materials.
Other Background
Operators reported that prior to December 1996, the data logger on stage 1 of the
hydrocracker experienced screen display problems.
At some point between December 1996 and January 1997,there were reports of stage 1
I/A system display problems which were being investigated.
On January 19, 1997 the unit was run while the I/A system for stage 2 was not
functioning optimally. There were gaps in the information displayed and at times
portions of the display would drop off.
During interviews two operators indicated that they knew of two instances when the
reactors were run above 800° F (an established high temperature that requires emergency
shutdown). One of these instances was reported on January 19, 1997: In both cases,the
operators took corrective actions and were able to control the temperature excursions and
the depressuring system was not used. It is also known from internal operating reports
that on July 23, 1992, a temperature excursion occurred and the 100 pound per minute
depressuring systems was used.
Operators were not aware that high off-scale values from the thermocouples default to a
value of zero. This caused confusion about the temperature readings during the few
minutes leading up to the explosion.
For the particular instrument configuration involved in this incident,high off-scale temperature readings
display a value of zero.
8 The rupture occurred in a straight run-no weld or elbow was involved in the actual rupture.
Page 6
CAUSES OF THE INCIDENT
Based on an analysis of information and interviews noted above, HSD has determined a
number of causes of this incident. For the purpose of this report, initiating cause,
contributing cause, root cause and cognitive tunnel vision as defined by the Center for
Chemical Process Safety (CCPS) are listed below:
Initiating Cause -the act, condition, or failure that resulted in the incident.
Contributing Cause -physical conditions, management practices, etc. that facilitated
the occurrence of an incident.
Root Cause -the prime reasons that allow faulty design, inadequate training, or
improper changes, which lead to an unsafe act or condition that results in an incident.
(If root causes were removed,the particular incident would not have occurred.)
Cognitive Tunnel Vision- a characteristic of human performance under stress.
Information is sought that confirms the initial hypothesis about the state of the
process while disregarding information that contradicts the hypothesis.
The Initiating Cause of the incident was the rupture of a 12 inch diameter pipe due to
over-temperature. The failure occurred on the outlet of the stage 2 number 3 reactor pipe
just upstream of the effluent feed heat exchangers.
The Contributing Causes of the incident are:
the reaction did not behave as operators expected;
it appeared to operators that the lack of hydrogen consumption and
loss of make-up hydrogen did not correlate with normal hydrogen
consumption associated with reactions at high temperatures;
- the operators had developed"cognitive tunnel vision"as they had
previously seen erratic temperature readings similar to those seen on
the evening of the incident,which had been caused by a temperature
indication problem;
operators were not aware that high off-scale temperature readings
would display a value of zero;
Page 7
- the temperature runaway occurred in about 7 minutes, which was
much faster than any other temperature excursions experienced at the
Tosco hydrocracker;
- the emergency situation was not apparent to the operators; and
established operating procedures were not followed regarding the high
temperature emergency shutdown requirement of the unit.
The Root Causes of the incident are:
a hot spot in bed 4 created a temperature runaway; and
the lack of training to provide the operations group with the understanding of
the need for prompt reaction to any unusual or conflicting temperature
indication.
DISCUSSION
Contributing Causes
Operators did not see hydrogen consumption data that they would normally expect to see
for high temperatures or for a temperature excursion. Typically, an increase in reactivity
for the catalyst would require additional make-up hydrogen. The hydrogen make-up flow
did not increase; in fact the hydrogen make-up flow decreased and by 7:37 p.m. the
hydrogen make-up dropped to zero. This was likely due to the fact that reactions at
abnormally high temperatures can create methane gas,which is heavier than hydrogen.
The heavier methane gases mixed with the hydrogen in the recycle gas causes an increase
in the density of the recycle gas. This causes an increase in pressure in the hydrogen
recycle system, which would cause the control system to decrease the hydrogen make-up
flow. Operators verified that the hydrogen purity analyzer showed on-specification
during the 7 minutes of the incident. There was confusion over the cause of the loss of
hydrogen make-up flow. The operators may have not been aware that the hydrogen
recycle purity analyzer has a 7 minute delay in reporting. By the time the analyzer
alarmed, the explosion had occurred.
From interviews, the operators indicated that it appeared that the data logger temperature
indication was malfunctioning. In the past, operators had seen the stage 1 data logger
display"lock up". According to the operators it was difficult to determine that there was
a problem with the display until it was noticed that the temperatures had not changed in
response to a control move. Tosco tested the data logger after the incident to try to force
Page 8
it to malfunction. Tosco was unable to replicate failures or malfunctioning of the data
logger.
Operators were not aware that a high off-scale temperature would display a value of zero.
This caused confusion with the temperature readings and lack of confidence in the data
logger during the few minutes leading up to the explosion.
The emergency situation was not initially apparent to the operators. This was likely due
to the erratic temperature readings displayed on the data logger and what appeared to the
operators to be a lack of correlation between the operating conditions and temperatures.
Root Causes
A hot spot in bed 4 of the reactor caused a temperature runaway. Possible causes for the
hot spot may have been poor distribution or a pocket of catalyst that spontaneously
developed a high reaction rate. It is possible that other unknown factors contributed to
the reactor hot spot.
Operators were trained in the shutdown of the unit but may not have been trained to
shutdown the unit in response to alarms or temperature indications that appear erroneous
or erratic. The operators did not know that high off-scale temperature values default to a
reading of zero. These items, combined with what the operators expected from past
operating experience with temperature excursions, likely caused general distrust that the
readings were accurate. Also, it appears that the operators had been able to control past
temperature excursions over 800° F without depressuring.
Post Incident Testing Results
Thermocouples were tested by the manufacturer. All thermocouples tested to be
accurate.
Catalyst testing by an outside firm verifies the temperature rise in beds 4 and 5. There
was a formation of mounds of catalyst in some of the beds. There were no indications of
other adverse catalyst conditions.
The section of pipe that failed was tested by an outside lab. The lab confirmed that the
point of failure on the section of pipe was not on a weld, elbow, or reducer. The lab also
Page 9
confirmed that the pipe failed due to over-temperature. The metallurgy and thickness of
the pipe were within design specification.
Other Possible Causes
During the course of the investigation a number of possible causes were investigated.
Possible causes were ruled out as appropriate evidence was gathered. The following
possible causes were considered and ruled out:
a control bounce occurred such as a large adjustment to temperature or
pressure;
loss of recycle hydrogen;
catalyst tray collapse;
de-activated catalyst was suddenly re-activated; and
temperatures in the reactor were too high for the feed rate.
RECOMMENDATIONS
HSD has recommended that the following items be implemented at the Tosco Avon
Refinery prior to start-up of the hydrocracker:
automatic high temperature shutdowns;
automatic high temperature deviation shutdown and/or alarms;
retraining of operators on the shutdown system,the reasons for the shutdown
requirement and when it is to be used;
retraining of the operators on the reasons for the critical operating variables;
retraining of the operators on the reaction kinetics of temperature excursions;
retraining of the operators on instrumentation default values; and
update of the operating procedures with a description of the instrumentation
default values.
HSD has also requested that Tosco share the findings and conclusions of their incident
investigation with other refineries and with their catalyst manufacturer and supplier.
CONCLUSION
All incidents, no matter where or how they occur, should be viewed as opportunities to
improve management systems rather than as opportunities to assign blame to individuals.
Page 10
The particular incident discussed in this report was not a simple failure but a complex
interaction of a wide range of contributing and root causes. These causes attest to the
importance of implementation of safety management systems, procedures, and policies
such as process hazards analysis, management of change, and provisions of proper
equipment and training for operations personnel.
Each of the contributing causes found in this incident has at its root a system failure. It is
important that all facilities embrace the concept of root cause analysis. Use of root cause
analysis for major incidents will result in a reduction of incidents with the same or similar
system failures.
Without addressing the issues identified as contributing and root causes to this event in a
proactive fashion, incidents such as the one that occurred on January 21, 1997 at the
Tosco Avon Refinery will continue to occur.
Page 11
ryk a
r
TOSCO REFINING COMPANY
REPORT ON THE JANUARY 21, 1997
HYDROCRACKER INCIDENT AT THE AVON REFINERY
' .
'
�
TABLE OF CONTENTS
I. Incident Summary.......... ....................................... ............................................. l
2I. Investigation Procedure.......................................................................................3
III' Hydrocracker Design
' and Operation ................................ ...................................5
A. -------------------------' 5
6
B. Stage and ----------------------.
C. Stage 2 Controls:.......... .......................... .............................. ..................9
D. Stage 2Reactor Temperature DinnlmyouozdAlarozS-----------. ll
T\/. Historical Operations ---.-------.----------.----.—.--' l4
A. Process ----------------------------' 14
B. Control -------.------'...-----.----. l5
l7
C. 1996 Run ------___..................................
'
lg
V. and —.—'--''''''--''—'-----''----'----'
�
VI. Events Preceding
' The Incident........................................................................... 2l '
��
V11. -----'.-----..-----.----...—.--.-----.------
-
3�
VI. Conclusions........................................................................................................
i'
I. Incident Summary
At approximately 7:41 pm on January 21, 1997, a pipe ruptured at the Tosco
Avon Refinery Hydrocracker Unit which caused an explosion and fire. This incident
caused the fatality of one Tosco employee, a hydrocracker operator named Michael
Glanzman, and forty-six minor injuries which were reported by Tosco employees and
contractors in the days following the incident. No significant off-site impacts were
reported.'
The pipe ruptured as a result of being subjected to excessively high temperature
and not being properly depressured, per emergency procedures. Cause for the excessive
temperature was a very rapid high temperature excursion' which started in catalyst bed 4
of Reactor 3 in Stage 2 of the hydrocracker. Although no single or conclusive cause for
the temperature excursion could be determined after the explosion, all possible causes for
the temperature excursion were identified. Possible reasons identified for the temperature
excursion include: spontaneous formation of a "hot spot" in bed 4; the flow
characteristics of fluids in bed 4; or the possibility of uneven liquid distribution within the
bed.
1 Details of the emergency response and notifications at the time of the incident are contained in the 72-
hour report made by Tosco to the Contra Costa County Department of Health Services,attached to this
report as Attachment A.
Z A temperature excursion is an episode in which temperatures deviate outside of a normal operating
range.
(1)
A contributing cause for the pipe rupture was that the operators did not depressure
the unit as specified in emergency procedures for safely controlling temperature
excursions. Reasons for the operators not depressuring the unit include.. the incident
occurred very rapidly and was outside previous experiences; they did not believe the
control room temperature indicators; and they were confused and distracted by process
information which was inconsistent with past temperature excursion experiences.
Corrective actions were identified to respond to the possible causes for the reactor
temperature excursion. These actions include installation of new controls which will
automatically depressure the unit in the event of a future high temperature excursion, and
retraining of operations personnel. This combination of hydrocracker corrective actions
will eliminate a future incident.
This report summarizes the results of a 3 month refinery investigation into the
causes of this incident. The first section of the report describes the investigation
procedure. The next several sections describe the physical design and operation of the
hydrocracker unit, as well as its control system, operational history, operator staffing and
training, and the events leading up to the incident itself. The last section of the report
presents conclusions and recommendations based on the investigation results.
(2)
_1
II. Investigation Procedure
Immediately following the incident, and in accordance with Tosco's incident
investigation procedures, the refinery manager directed that an investigation committee be
formed, comprised of the following staff members:
• Engineering Manager
• Technology Manager
• Mechanical/Metallurgical Engineer
• Field Safety Supervisor
• OCAW Health and Safety Representative
• Health and Safety Manager
• Refinery Loss Coordinator
• Air Toxics Coordinator
• Senior Process Engineer
The incident investigation included gathering information available from the scene
based on visual observations, detailed inspections, collection and analysis of strip charts
and digital information, collection and analysis of catalyst and metallurgical samples,
discussions -with eye witnesses, collection and review of drawings, procedures and
historical information on the operation, monitoring and maintenance of the unit, as well as
(3)
a technological review of the hydrocracking process. Thermocouples and temperature
monitoring equipment were tested to ascertain their functionality and thereby determine
their performance during the incident. The data was analyzed to develop an exact and
consistent timeline of events. Once the sequence of events was established, the unit
process, its control systems and the human factors associated with the operation were
considered. This led to identification of factors which might have contributed to the event
as well as development of recommendations to prevent a recurrence. The investigation
team met daily for over 2 1/2 months to review their findings .as the investigation
proceeded.
In addition, a number of regulatory agencies participated in investigations of the
incident, under the leadership of Cal/OSHA. These agencies included federal OSHA, the
Contra Costa County Health Services Department, U.S. E.P.A., and the Bay Area Air
Quality Management District. Tosco provided temporary on-site office facilities for
representatives of these agencies who remained on-site for over 2 1/2 months. Tosco also
provided thousands of pages of documents for review, as well as responses to numerous
written and oral questions, tours, and briefings. As lead agency, CaUOSHA conducted
witness interviews based in part on questions provided by representatives of the other
agencies, and Cal/OSHA briefed those agencies on the interview results. In addition,
Tosco conducted several meetings to brief county staff on the investigation results and
respond to County questions, and arranged for county staff to interview.employees
concerning remaining questions and issues requiring further clarification.
(4)
III. Hydrocracker Design and Operation
A. Hydrocracker Complex:
The hydrocracker complex consists of a hydrogen plant and the first and second
stages of a hydrocracking unit. Feedstocks for the hydrocracker include gas oils from the
crude units, vacuum units, Coker, Fluid Catalytic Cracker (FCCU), and hydrogen. Stage
1 of the unit hydrotreats the feedstock to remove nitrogen and sulfur, utilizing three
parallel reactors designated as Reactors A, B, and C. Stage 2 of the unit hydrocracks the
treated gas oil produced by Stage 1, using three parallel reactors designated as Reactors 1,
2 and 3. Products from the hydrocracker include depropanizer overhead (which can be
used as hydrogen plant feed); a mixture of iso and normal butane; light hydrocrackate
(which is a gasoline blend stock); heavy hydrocrackate(which is fed to the reformers); and
splitter bottoms (which can be blended to, diesel or sent to a downstream unit for
saturation of aromatics).
(5)
B. Stage 2 Design and Operation:
Stage 2 of the hydrocracker consists of high pressure and low pressure systems.
The hydrocracking reaction occurs in the high pressure system,' which operates in the
range of 1350 to 1735 psig. The high pressure system includes charge pumps,
feed/effluent heat exchangers, trim furnaces, reactors, product coolers, a high pressure
separator and a recycle compressor. The section of piping which ruptured during the
incident was a portion of the effluent piping -- a 12" diameter pipe running from the
bottom of Reactor 3 to the top of a 40 foot high heat exchanger structure. The pipe
ruptured adjacent to a 12" x 10" reducer in a straight run, horizontal section of the line
just prior to where it enters the top exchanger. There was not any weld failure involved.
The low- pressure system separates the reaction products for blending and
downstream processing in the refinery. Major equipment in the low pressure-system
includes a low pressure separator, as well as stabilizer, depropanizer and splitter towers.
A portion of the bottoms stream from the splitter tower is recycled to the high pressure
system.
Within the high pressure system, hydrogen and gas oil, feed are pressurized,
combined,.heated, sent to the reactors, catalytically converted, and then cooled. Gas oil
feed, consisting of Stage I product and recycled splitter bottoms, is fed to each of three
See Figure 1,a process flow diagram of the high pressure system,
(6)
parallel reactors and the associated bank of feed/effluent exchangers at an individually
controlled rate. The oil feed to each reactor is combined with preheated -hydrogen and
heated in the respective feed/effluent exchangers by heat exchange with the reactor
effluent. The feed is then brought to the desired reactor inlet temperature of 600° to 6541
by the addition of hot hydrogen from the trim furnaces. The heated mixture of oil and
hydrogen enters the top of each reactor.
Within each reactor are five beds of catalyst, through which the feed passes
sequentially. Cool hydrogen is distributed uniformly across the reactor though a
perforated pipe distributor and is injected into quench zones above the 2nd, 3rd, 4th and
5th beds to control the rate of the exothermic (heat producing) hydrocracking reaction.
The hydrogen as well as hydrocarbon liquids and vapors are collected on quench trays
above the 2nd, 3rd, 4th and 5th catalyst beds, and mixed in a quench box in the center of
each tray. The mixture is redistributed to the bed below by passing over and through
distribution trays. There are two distribution trays above the 2nd, 3rd and 5th beds, and
one distribution tray above the 4th bed in each reactor.
Reactor effluent from the 5th and final bed leaves each reactor through the reactor
effluent line. The reactor effluent line leads to the feed/effluent exchangers where the
effluent is cooled through heat exchange with reactor feed. The cooled product from each
set of feed/effluent exchangers combines with streams from the parallel reactors and
(7)
exchangers. The combined product is cooled further by exchanging heat to preheat
hydrogen and stabilizer feed, by air cooling in the reactor effluent air coolers, and by water
cooling in the reactor effluent condensers.
The reactor temperatures needed to maintain the desired level of conversion
depend upon the quality of the feedstock, the desired level of conversion and the activity
of the catalyst. The reactor temperatures are controlled through management of the trim
furnaces and the quench flows. The activity of the catalyst generally declines over time.
Catalyst life ends when it is no longer possible to maintain the desired level of conversion
without exceeding temperature constraints. For the hydrocracker, this includes a
maximum outlet temperature of 690° F, a maximum temperature rise across any catalyst
bed of 45° F or a maximum radial temperature difference in any bed of 45°F. Catalyst life
for the Stage 2 catalyst is typically 4 years.
The cooled reactor effluent is separated into a liquid and vapor phase in the high
pressure separator. The liquid phase is let down in pressure to the low pressure section of
the second stage. The recycle compressor recompresses the vapor phase, a mixture of
hydrogen and light hydrocarbon gases. Hydrogen partial pressure, which is the
mathematical result of hydrogen purity and unit pressure, is targeted at a minimum of
1100 psia to preventcatalyst deactivation due to petroleum coke build up. The purity and
pressure of the recycle stream are monitored to maintain this target. If the purity of the
(8)
recycle stream is too low, a portion of this stream can be bled from the system to purge
light hydrocarbon gases. The recycle stream is utilized as quench hydrogen or heated and
combined with the oil feed. Make up (supplemental) hydrogen is added to maintain the
pressure in the recycle system.
C. Stage 2 Controls:
(See Process Flow Diagram attached as Attachment A) The FC 93 oil control
valves set the rate of feed to Stage 2. The PC 120 valve regulates a recycle stream to a
section of the make up compressors so that a sufficient quantity-of hydrogen is added to
maintain the pressure in the recycle hydrogen system at the discharge of the recycle
compressor. The FC 97 hydrogen flow control valves are set to maintain a sufficient
hydrogen to oil ratio. Hydrogen flow to the trim furnaces is adjusted by the FC 109
valves. TC 125, which controls the reactor inlet temperature, sends a signal to TC 124 to
increase or decrease the trim furnace outlet temperature as required to maintain the
desired reactor inlet temperature. TC 124 sends a corresponding signal to PC 121 to
increase or decrease the fuel gas pressure, and thus the firing rate, to the furnace. The
inlet temperature to each of the lower 4 beds in each reactor is controlled by the TC 137,
138, 139, and 140 control loops. These loops control the flow of hydrogen to the quench
zone above each of these beds to maintain the desired bed inlet temperatures.
(9)
An on-stream analyzer is used to monitor the hydrogen purity in the Stage I and
Stage 2 high pressure systems. The analyzer is cycled between gas samples from the two
stages. High temperature reactions tend to produce methane and other light hydrocarbons
which are found in natural gas. Since these light hydrocarbons accumulate in the high
pressure system, a drop in the hydrogen purity is one indication of a temperature
excursion. A post-incident study of the analyzer and sampling system determined that the
time required for the analyzer to indicate a change in the process was approximately 7
minutes.
The unit is also equipped with depressuring systems. One system, which
depressures the unit at 100 psi per minute, activates automatically upon loss of the recycle
compressor. The other system, which depressures the unit at 300 psi per minute, must be
manually activated. When activated, the 300 psi system opens a valve which depressures
the high pressure system to the refinery flare, diverts make up hydrogen, and shuts down
the Stage 2 charge pumps, the firing to the trim furnace, and the recycle compressor.
Stage 2 reactor temperatures are also electronically monitored, though not
controlled, by 96 array thermocouples in each of the reactors, as well as thermocouples in
each of the feed and effluent lines, and three thermocouples attached to the exterior shell
of each reactor. Of the 96 thermocouples s in the catalyst beds, twelve are located in the
top bed, twenty-four are located in each of the next three beds, and twelve are located in
the lower bed. The orientation of the array thermocouples is shown on Figure 2. These
(10)
thermocouples are used to monitor the temperature in the beds to determine the axial
temperature gradient (the temperature difference between points above and below each
other in a catalyst bed), and to determine the radial temperature gradient (the difference in
temperature among points at the same level in the catalyst bed).
D. Stage 2 Reactor Temperature Displays and Alarms
There are two sets of temperature displays in the hydrocracker control room for
the Stage 2 reactors: control board mounted instruments, and a PC based "data logger"
display. As described below, the board mounted instruments and the data logger are
connected to some of the same thermocouples located at the inlet and outlet of each bed.
They therefore function as redundant systems in terms.of providing information on the
temperatures in each bed.
The control board instruments (also known as "Moore controllers") display
information in digital, LED light bar, and strip chart format. These instruments show the
following temperature and flow information for each reactor: hydrogeq flow,to the trim
furnace; .trim furnace outlet temperature; reactor: inlet temperature; 1st bed outlet
temperature; the. difference between the reactor inlet and outlet temperature; inlet
temperature, outlet temperature and quench flow for the 2nd, 3rd and 4th catalyst beds;
5th bed inlet temperature and quench flow; reactor outlet temperature; hydrogen flow to
the feed effluent exchangers; oil feed rate; and. fuel gas flow to the furnaces. For
monitoring purposes, the control board temperature displays are linked to single
thermocouple points at the center of the inlet and outlet of each bed. However, the
quench controls are linked only to the inlet temperature points of each bed, and do not
control on the basis of outlet temperatures.
The data logger, consisting of a computer with associated interfaces and software,
is linked to 40 of the 96 array thermocouples described above, including the center
thermocouples connected to the board mounted displays as well. The remaining 56
thermocouples are linked to external panels located near the bottom of each reactor. The
data logger displays five temperatures at the outlet of the first catalyst bed, five
temperatures at both the inlet and outlet of the next three catalyst beds, and one
temperature at the inlet and four at the outlet of the fifth bed: The data logger also
displays the three reactor external wall temperatures, and reactor inlet and outlet
temperatures. It also calculates and displays averaged bed temperatures. The data logger
receives a digital signal, over a single pair of wires, from a field multiplexer which is
connected to the thermocouples. The range of the multiplexer is 0° F to 1400° F.-.If a
thermocouple fails or.indicates a temperature that is out of range, the multiplexer sends a
0 signal to the data logger. The data logger displays updated temperatures at 15 to 40
second intervals.4
The data logger is also programmed to retain a record of temperature indications,.known as the data
logger historian The historian program includes a data compression algorithm that uses exception
reporting of changing readings. On a periodic cycle of one hour, the historian records the current value
for all points. Between these periodic readings, the points are recorded and timestamped only if they
change more than a predetermined threshold(deadbarid).
(12)
The control room contains the following alarms related to reactor operation: a
flashing signal from any of the quench controllers if the quench valve position is more than
50% open on beds 2, 3 and 5 and 75% on bed 4, and a data logger alarm for bed
temperatures above 780° F. The data logger alarm consists of a flashing light and audible
horn on the second stage alarm panel as well as a blinking off color display of the
temperature point on the data logger screen.
The data logger has one digital output for high temperature that is connected to
the Stage 2 alarm panel. Once the alarm is activated, the operator must acknowledge the
board annunciator to silence the horn and stop the light flash. The operator must then
acknowledge the data logger to turn off the light and prepare the system to alarm again
when another high temperature occurs.
Printouts from the historian will print a value for each point at each time interval requested. If no
data was recorded for a point at a time interval,the historian fills in the value from the last previous
recorded value;the program assumes that the temperature has not changed significantly(more than the
deadband)from the previous value. The historian creates a new data file every eight hours. These files
are stored on the computer hard drive for one month.
(13)
IV. Historical Operations
A. Process History:
The unit was originally started up in 1963 under a license from Chevron Research
Corporation. In 1986, the design was modified in accordance with a technology license
from Union Oil of California. Modifications at that time included addition of a
depressuring system, new reactor thermocouples, new hydrogen quench rings for Stage 2,
installation of reactor internals in each of the Stage 2 reactors and installation of single
loop digital controls ("Moore Controllers") for the hydrogen quench system.
The unit is scheduled for major maintenance turnarounds every two to three years,
typically limited by catalyst run life and feed effluent exchanger performance. The last
major turnaround of the unit was completed in January 1996. Typical catalyst life is two
years for the Stage 1 catalyst, four years for the Stage 2 catalyst and three to five years for
the hydrogen plant catalyst. Work during the,1996 turnaround included replacement of
the Stage 1 catalyst, installation of fresh catalyst in the top beds of Stage 2 and installation
of regenerated catalyst in the other beds, modification of the top bed quench distributor
trays in all reactors, and installation of the 96 array thermocouples in the Stage 2 reactors
to provide more temperature data for each catalyst bed. Thirty-six companies were
employed-to complete-field work during the hydrocracker turnaround.
(14)
B. Control System History:
The original controls for the unit were Honeywell analog electronic controllers and
recorders, with a Metroscope used for temperature monitoring. Over time, a number of
control modifications were made to improve reliability, accuracy and precision in the
control system. In 1978, Stage 2 feed controls were changed to Foxboro Spec200, a set
of analog electronic controllers. In 1986, the Stage 2 reactor quench controls were
changed to Moore 352 single loop digital type controllers. In 1986, a Foxboro Videospec
DCS (Distributed Control System) was installed on Stage 1 and the hydrogen plant. In
1989, the low pressure section of Stage 2 was converted to Moore 352 single loop digital
type controllers and the PC based data loggers for Stage 1 and Stage 2 were installed.
The data loggers offered improved alarm and display capability and replaced a device (the
Metroscope) that had become difficult to service due to changes in technology and the
unavailability of replacement parts.
In 1990, the Videospec DCS system for Stage 1 and the hydrogen plant was
replaced with a Foxboro I/A (Intelligent Automation) DCS system. The Foxboro UA
system is the current Avon refinery standard for distributed digital control systems. The
I/A system consists of Operator Work Stations, Control. Processors, and Field Bus
Modules (FBM's). - The Operator .Werk Stations display. process data and accept
operating commands from the operators. The Control Processors contain control logic,
and.receive input from the workstations and FBM's. The control processors transmit data
(15)
to the workstations and transmit data and commands to the FBM's. The FBM's provide
the field interface between the field instrumentation and the control processors. The UA
system has numerous features to insure its reliability which include a protocol for
redundancy in the control processors and operator workstations as well as a hardened and
sealed environment within the FBM's.
In 1994, the Stage 1 UA system was upgraded by providing additional consoles
and alarm displays. In July 1995 a project, which was completed in January 1996, was
initiated to install 96•array type thermocouples in each of the Stage 2 reactors to provide
more temperature information. In December 1995, a project was initiated to replace the
Stage 1 and Stage 2 data loggers with temperature displays of all the array thermocouples
linked to the UA system.
On December 24, 1996, the Stage 1 monitoring points were brought into the UA
system. On January 10, 1997, the 40 temperature monitoring points displayed on the
Stage 2 data logger were transferred from the data logger to the UA system. The other 56
points remained connected to the external temperature monitoring panels on each reactor.
The Stage 2 UA system operated for two days, before the unit was shutdown due to an
internal tube leak inside a Stage 1 heat exchanger. When the unit came on line again,
during the weekend of January 19, the operators reported that the Stage 2 UA system was
malfunctioning; specifically, the average bed temperatures were not being calculated
correctly, and certain temperature points were simply "dropping off' the screen. On
(16)
January 20, 1997, the Stage 2 points were removed from the UA system and returned to
the data logger, pending repairs to the UA system.
C. 1996 Run History:
Following the January 1996 turnaround, the hydrocracker unit was started up on
February 16, 1996. On March 19, 1996, there was a temperature excursion in Stage 2,
Reactor 1. The excursion began in bed 3 and progressed to beds 4 and 5. The maximum
estimated reactor outlet temperature during the excursion was 920'. The operators pulled
oil feed from Reactor I about 3 minutes after the bed 4 outlet temperature exceeded 800°
F. Within about 7 minutes of pulling feed, temperatures at the outlet of bed 4 began to
fall. Within another 6 minutes, the reactor outlet temperature also began to fall.
Management review of this incident resulted in two recommendations: (1) to replace the
Spec200 controllers used to control the oil feed to-the Stage 2 reactors with Moore 352
controllers; and (2) to review and reissue the guidelines for temperature control in the
Stage 2 reactors. The oil feed controller replacement was completed on June 28, 1996.
On April 4, 1996, the temperature control guidelines were reissued to all
operators. These guidelines specified that:
(17)
(1) Maximum axial or radial temperature differentials in a catalyst bed must be held to
less than 45° F.
(2) Bed inlet temperature must be reduced if any temperature rises 5° F above normal.
(3) Oil feed to a reactor must be pulled if any point is 25° F over normal.
(4) The unit must be depressured at 300 psi per minute if any point is 50° F above
normal or over 800° F.
Safety meetings were held on these procedures, and the guidelines were posted on
the control board for the Stage 2 reactors. The crew on shift during the January 21
incident again reviewed these procedures during a training session in November 1996.
Despite the fact that the procedures required depressuring the unit if any
temperature exceeded 800° F,-a review of the operating records indicates that, on four
occasions after the guidelines were issued, the operators continued to use quench or
pulled feed to control apparent temperature excursions, rather than depressuring the-unit
as required. Three of these events appear related to faulty temperature monitoring
equipment. In July 1996, the Stage 2 Reactor 3 outlet temperature signal to the control
board as well as the data logger display was lost, apparently due to a failed thermocouple.
In September 1996, the Stage 2 data logger failed to perform properly on two occasions.
Once, the data logger stopped updating twice and had' to be reset by instrument
technicians in order to restore service. On another occasion, it was reported that the data
(18)
logger had stopped working and repairs were made to restore it to service. Operators
relied on board mounted instruments to continue operating the unit. The fourth incident,
on January 19, 1997, is described in detail below. Many of the operators reported that
they have experienced numerous temperature excursions, but most could recall only one
instance when the unit was depressured as a result. One operator reported that he had
been told he could be fired for failing to depressure when necessary in the future.
Operators reported that excursions typically showed up as a "wave" passing sequentially
through the reactor beds from top to bottom, in a predictable pattern.
V. Operator Staffing and Training
There are normally five operators on duty at the hydrocracker during each of three
shifts (day, swing and night). One is a No. 1 operator and the others are No. 2 operators.
The No. 1 operator oversees the shift, assists with board duties if necessary and makes
outside rounds at least once per shift. The No. 2 operators have assigned duties as
follows:
Hydrogen board -- the hydrogen board operator operates the UA.control system
for the hydrogen plant and Stage 1.
(l9)
Iso board -- the iso board operator operates the control system for Stage 2,
including both the high and low pressure systems.
East pad -- the east pad operator is responsible for making rounds to check
equipment, taking outside readings and obtaining samples as necessary for the hydrogen
plant.
West pad -- the west pad operator is responsible for making rounds to check
equipment, taking outside readings and obtaining samples as necessary for the Stage 1 and
Stage 2 high and low pressure systems.
On the night of the incident, all operating positions were filled by qualified,
experienced personnel. In addition, two extra operators were on shift. One was an extra
No. I operator, held over from day shift to monitor repairs to an engineered clamp that
had been leaking hydrocarbons on the outside of a Stage 1 heat exchanger. The other was
a No. 2 operator held over to bring Reactor A up to temperature before pressurizing it as
scheduled for day shift on January 22.
Training of operators at the hydrocracker includes on the job training (consisting.
of job shadowing and instruction by other qualified operators),and a combination of oral,
field demonstration or written tests, administered by the No. 1 operators and unit
supervisors. The average training time to qualify a hydrocracker operator is 18 months.
(20)
VI. Events Preceding The Incident
On January 12, the hydrocracker unit was shutdown for repairs to an internal tube
leak inside the Stage I stripper feed preheat exchanger. On January 17, the unit was
started up again. Oil was introduced to Stage I at 4:25 am. Feed in to Stage 2 was
delayed slightly due to the need to fix a hydrocarbon leak in an engineered clamp between
two of the Stage I feed/effluent exchangers. Feed was introduced to Stage 2 at 5:40 pm
on January 18. By approximately 10:00 pm on January 19th, Stage 2 products were on
specification.
During the night shift on January 19th, the operators experienced a rapid, sudden
temperature rise in bed 4 of Reactor 1. According to the plant information ('Pr') system,
the center outlet temperature in that bed increased from 653° F to over 800° F during the
20 minute period from 10:20 to 10:40 pm. The Iso Board Operator reported seeing .
temperatures on the control board increase to 660° F - 670° F. -He opened the hydrogen
quench valves to beds 3 and 4. Bed inlet temperatures came down, but the bed 4 center
outlet temperature continued to increase to more than 800° F as indicated on the control
board display. The West Pad man went outside to check the external temperature panel
and reported temperatures in excess of 904° F. The No. 1.operator directed that feed be
pulled from Reactor I and fuel gas flow to the trim furnace be reduced. According to PI
information, the bed 4 center outlet temperature reached a maximum of 998° F about five
(21)
minutes after pulling the oil feed. It then decreased, falling below 800° F in about I
minute. The operators then continued lowering Reactor I temperatures to 550° F, and
reintroduced feed approximately one hour later.
During the January 19 incident, the UA temperature monitoring system for Stage 2
was in service, and the 40 bed temperature points from the data logger had been tied into
the system. The other 56 points remained connected to the external monitoring panel.
Because only 40 of the points.were connected, average bed temperature calculations by
the UA system were incorrect. In addition, operators reported that over half of the
temperature points were periodically dropping to 0. It was determined that this was most
likely due to.software interface problems. On January 20, following discussions among
the operators, unit supervisors and instrument engineers, the•supervisors decided to
disconnect the monitoring points from the UA monitor and reinstall them on the data
logger, until such time as the UA problems could be solved.
By the middle of the night shift on January 20, 1997, Stage I rate had been
increased to 25,300 bpd, and temperatures and reactions in both stages were satisfactory.
At approximately 4:50 am on January 21, the clamp on the Stage 1 feed/effluent
.exchanger began lealdng again. Feed was pulled from Reactor A at 5:20 am and.oil flow
to Reactors B and C was increased. , Because temperatures in Reactors B and C
decreased, the nitrogen content of the Stage I product rose from 4.6 ppm at 4:00 am, to
352 ppm at 10:00 am. High nitrogen levels in the Stage 2 feed tend to reduce catalyst
(22)
activity. In this instance, as a result of the high nitrogen content of the Stage 1 product,
cracking in Stage 2 declined.
During the day shift on January 21, 1997, differential temperatures across the
Stage 2 catalyst beds averaged less than 10° F per bed and the unit was not producing any
light product, indicating a lack of hydrocracking reaction. At the direction of the
supervisors, the operators continued adjusting rates and temperatures in Reactors B and C
throughout the day, staying below the posted maximum reactor outlet temperature of 760°
F, but trying to increase the reaction and thereby lower the nitrogen level in the Stage 2
feed. The feedrate to each of the Stage 2 reactors was maintained at 6,000 bpd, consistent
with operating guidelines for minimum flow during.unit start up. At 5:40 pm, a sample
from the Stage 1 low pressure separator contained 9.5 ppm nitrogen, which is within the
normal range for Stage 2 feed.
-During the day shift on January 21, a contractor injected sealant into the leaking
engineered clamp on the Reactor A feed/effluent exchanger, in order to stop the leak.
Feed was to be kept out of Reactor A until the morning of January 22, to allow time for
the sealant to cure. The supervisors specified an operating plan for the evening of January
21 which required the operators to continue to raise temperatures in reactors B and C at a
reduced rate,in order to get-the nitrogen down to 5 ppm or less, and then to raise the rate
in Stage 1 if possible. In addition, the operators were directed to slowly and gradually
increase temperatures in Stage 2 in order to drive the nitrogen off the catalyst. The
(23)
objective for Stage 2 was to continue gradually increasing temperatures to restore
reactivity. This is a standard industry practice for dealing with nitrogen deactivated
catalyst, typically resulting in restored catalyst activity over the course of several days or
weeks.
At the start of the swing shift on January 21 (2:00 pm), there were no light
products in the low pressure section of Stage 2; indicating little or no reaction was
occurring. Stage 2 bed inlet temperatures varied from about 612° F to 640° F and were
steady. Quench flows (which are one indication of reactor activity) were low: only beds 2
& 3 in Reactor 1, bed 2 in Reactor 3, and bed 3 in Reactor 3 were above 10%.
At 7:34 pm, the data logger alarm sounded. The alarm was apparently triggered
by an increase in a single temperature point on the side of the outlet to Reactor 3, bed 45,
which rose from 628° F to 823° F in 40 seconds. At the same time, data logger records
indicate that the bed 5 inlet temperature rose from 637° F to 860° F. The operators
reported seeing data logger readings at about this time of 690° F on the 4th bed outlet,
and 890°.F on the 5th bed inlet. The control board indicated the bed 5 inlet temperature .
increased from about 640° F to full scale (800° F). The quench flow controller to bed 5
began to open in response to the increase in inlet temperature.
5 Unless otherwise noted,all further references to catalyst beds are Reactor 3 beds.
(24)
At approximately the same time period, the demand for make up hydrogen began
to fall. An increase in reactivity typically consumes hydrogen and therefore could be
expected to cause an increase in demand for hydrogen. However, it appears that the
reduced hydrogen demand was actually indicative of high temperature reactions and
resulting formation of light hydrocarbons in Stage .2. When light hydrocarbons are
produced as a reaction product, they build up in the recycle system. Since these
hydrocarbons are much heavier than hydrogen, the density of the recycle stream increases.
The higher density of the recycle stream causes an increase in pressure at the discharge of
the centrifugal recycle compressor. .This in turn causes the Stage 2 pressure controller to
reduce hydrogen make up which was contrary to what the operators stated that they
expected to see during a temperature. excursion. The operators were monitoring the
hydrogen analyzer, but it did not detect the drop in hydrogen purity due to the lag time
inherent in the design of the analyzer and its associated sampling system.
The Iso Board operator heard the data logger alarm, noted the high temperaLture in
bed-4 on the data logger, but noted that the control board strip chart for bed 4 looked
normal. This isconsistent with the fact that the temperature point which alarmed on the
data logger is not a point that is linked to the control board display. The operator
expressed concern over a potential excursion and within a minute the,No. I operator on
shift as well as the extra No. I joined hiih in evaluating the control board and data logger
readings. They then reported seeing the data logger temperatures start"bouncing up and
down," from normal range temperatures to 0 and back again.
(25)
By 7:35 pm, the data logger historian indicates the bed 4 outlet temperature
decreased to 637° F. The bed 5 inlet temperature also began to decrease. By 7:36 pm,
the quench flow to bed 5 had risen to full scale (24 MM SCFD) and the bed 5 inlet
temperature had decreased to 633° F as indicated by the data logger historian, as well as
the observations of the operators and the control board strip chart. However, by this time
the reactor 3 outlet temperature had increased from 641° F to 650° F, as indicated by the
data logger and by the strip charts for bed 5 and the reactor trim furnace. It appears that
none of the operators noticed this temperature increase, which was an indication that the
excursion had spread to the reactor outlet.
Throughout this time, the operators report that the temperatures on the data
logger continued to "bounce up and down," fluctuating between high, normal and 0
temperature readings. At some time prior to 7:37, Michael Glanzman, the West Pad man,
went outside to check the external temperature monitoring points near the reactor.
Between 7:36 and 7:37 pm, the fuel gas pressure at the Reactor 1 furnace had
increased to 30 psi vs. a posted maximum of 28 psi. The extra No. 1 operator reduced
firing in the Reactor 1 trim furnace to prevent overfiring, but then became concerned
about losing temperature in the reactor system, in response to what he saw as a decreased
(26)
Reactor 3 bed 5 inlet temperature. He therefore manually reduced the Reactor 3, bed 5
quench flow.
Between 7:37 pin and 7:39 pm, the extra No. I operator worked to control the
trim furnace firing, while the Iso Board man and the No. I operator continued to monitor
the temperatures on the data logger, which reportedly continued to fluctuate and/or drop
to 0.
At 7:37 pm, the hydrogen makeup fell to 0. The Hydrogen Board man called out
that make up had been lost. The Iso Board man checked the control board, noted that the
quench flow to bed 5 was manually closed and, at 7:38 pm, opened it back up.
Between 7:38 pin and 7:39 pm, data logger records indicate that each of the four
temperatures at the outlet of bed 5 rose to more than 780° F, with one of the points
indicating more than 1200° F. At 7:39 pm, three of the data logger readings defaulted to
0 and one point was reading 889° F. The data logger indicated that the reactor outlet
temperature rose to 983° F. The control board indicated that the reactor outlet
temperature rose to full scale (800° F). Both the control board and the data logger also
indicated an increase in reactor inlet temperature.
Between 7:39 pm and 7:40 pin, the data logger records indicate that the reactor
outlet temperature increased to 1220° F and then defaulted to 0. The control board
(27)
display of reactor outlet temperature continued to read off scale high. The control board
display of reactor inlet temperature rose to full scale (800° F) and the data logger records
indicate the reactor inlet temperature rose to 889° F. At about this time, the operators
heard a radio.message regarding temperatures from Michael Glanzman that was garbled
and unclear. Two of the operators went outside to check on him.
After 7:40 pm, the control board indications of reactor inlet and outlet temperature
continued to read off scale high. The data logger records indicate the reactor inlet
temperature reached 1234° F before defaulting to 0. Data logger indications of two of the
bed 5 outlet temperatures changed from their default readings. One, on the south side of
the reactor indicated 880° F and then decreased to 695° F. The other, on the northwest
side of the reactor, indicated 1398°F.
At about this time, the extra No. I operator called the shift supervisor by radio and
asked him to call back by phone. The supervisor immediately returned the call, and the
operator requested the assistance of an instrument technician and indicated there. were
problems with the unit. At about the same time, the Iso Board man noticed the control
board indications of high,temperatures atthetop of the reactor. In response, he reduced
firing on the trim furnace and lowered the temperature set points to the top two beds, as a
means of increasing quench flow. At 7:41, shortly after the extra No. I operator hung up
the phone, while the Iso board man was lowering the set point on bed 3, and when the two
other operators had just exited the control room, the explosion occurred.
(29)
Observers report a pop or crack, followed by a whoosh, followed by a boom. One
observer reports seeing a section of the pipe glowing red, seconds before the explosion.
Based upon these observations, it appears that the hydrocarbons auto-ignited very shortly
6
after the initial release.
VII. Post-Incident
Physical analysis:
Following the event, there was a visible rupture of the effluent piping out of
Reactor 3, on a straight run of pipe, not at a weld. The piping was originally specified as
12 inch schedule 100 piping constructed of 1-1/4 % Chromium 1/2% Molybdenum steel.
The pipe is believed to be from the original unit construction in 1963. Original wall
thickness was specified to be a minimum 0..749 inch. A 1991 ultrasonic inspection in the
near vicinity of the failure showed a thickness of 0.94 inch. A 1995 inspection of the
reactor 3 effluent piping showed no significant indication of thinning. Preliminary analysis
of the failed section indicates that the pipe, at the point of failure, had expanded in
circumference by approximately 5 inches,,which created a localized bulge in the pipe prior
'Emergency response activities after the explosion are described in the 30 day report to the County
Department of Health Services,attached hereto.
(29)
to the rupture. Other sections of the reactor 3 effluent piping had also expanded. This is
consistent with the short term creep of 1-1/4% Cr 1/2% Mo steel at temperatures above
1300° F, with a nominal 0.90 inch wall thickness, and an internal 1400 psig to 1500 psig
pressure. The composition of the piping was confirmed to be within the specifications for
1-1/4% Cr 1/2% Mo pipe.
A detailed inspection of all three Stage 2 reactors was conducted after the incident,
and catalyst samples were collected and analyzed. The quench tray above the 4th bed in
each reactor appeared to have been disturbed as a result of the incident, and pillars of
agglomerated catalyst were found in each of the 4th beds.- In Reactor 3, about one-fourth
of the tray was bent downward, 12 inches below the tray support ring, and the manway
was lifted 12 inches above the adjacent tray sections. The upper surface of the bed 4
catalyst appeared to have a 2 foot depression in the center. . A pillar .of agglomerated
catalyst was found, which extended 8 feet upward from-the support grid at the bottom of
the catalyst bed., The pillar was located in the center of the catalyst bed, was about 2 feet
in diameter at its base, and about 8 inches in diameter at the top. Catalyst samples from
the pillar (as well as the pillars in the other two reactor 4th beds) were 2.5 to 3 times
higher in carbon content due to the buildup of petroleum coke than samples of loose
catalyst from the same beds. In Reactor 3, the quench tray manway above bed 5 was
warped about 3 inches above adjacent tray sections. Again, this may have preceded the
incident or could have been a result of the dramatic pressure decrease during the incident.
(30)
Catalyst samples from the middle of bed 5 showed an 80% loss of surface area, consistent
with exposure to extreme temperatures.
Temperature Monitoring:
Post-incident analysis indicates all the thermocouples and the data logger were
functional at the time of the incident. The data logger historian indicates a series of
temperatures that were entirely consistent with the strip chart record when both the strip
charts and the data logger were within their respective temperature ranges. It is clear
from both the operators' statements and the data logger historian that the control panel
alarm, initiated by the data logger, sounded at approximately 7:34 pm. Subsequently, the
reported "bouncing" of the data logger temperatures may be explained by the fact that the
data logger displays 0 when temperatures exceed 1400° F, its maximum range, and it
averages these temperatures as 0 when it calculates average temperatures. It is not
possible, however, to verify independently,what was displayed on the data logger screen at
the time of the incident.
Responses to the 1ERh Temperature Excursion:
Post-incident analysis indicates thit, at the time the alarm sounded at 7:34 pm,
neither quenching nor pulling feed would have been sufficient to control this temperature
(3 1)
excursion. The recorded inlet temperature to bed 5 was 860° F at 7:34 pm. Reaction
kinetics indicate that quench would not have provided sufficient cooling to reverse the
reaction. It also appears that pulling feed would have been ineffective to stop this
excursion. During the prior incident on March 19, 1996, the reactor outlet temperature
did not begin to decrease until 13 minutes after pulling feed. On January 21, the entire
incident lasted 7 1/2 minutes, from onset of the excursion to pipe rupture and explosion.
Therefore, even if feed had been pulled at the onset of the excursion (7:34 pm), it appears
.the effect would have been too slow to prevent the rupture.
VI. Conclusions
The incident appears to have been caused primarily by two factors:
I There was a very rapid temperature excursion in the lower beds of Reactor 3,
which appears to be related to localized temperature instability in the 4th bed, the
cause of which cannot be determined.
2. In response to the temperature excursion, due to a variety of factors, the operators
did not depressure the unit as required by the emergency procedures.
(32)
The High Temperature Excursion
The entire duration of this incident was 7 1/2 minutes. In that short time, a highly
localized high temperature spot in bed 4 of the reactor initiated a rapid temperature
excursion in bed 5 that traveled through the reactor outlet and caused a rupture of the
outlet piping. This happened despite the fact that the temperature had been reduced to
near normal levels in the lower 4th and upper 5th beds before the rupture occurred. Such
an excursion is unprecedented in the unit's 34-year history of operation.
. The rapid temperature excursion appears to be related to localized process
.instability.7 The cause of that instability, however, cannot be conclusively determined.
Several factors may have played a role: the exothermic nature of the
hydrocracking reaction which can cause,spontaneous evolution of"hot spots;" the flow
regime in bed 4; or the possibility of uneven liquid distribution within the bed.
Other possible explanations were ruled out for a variety of reasons. Reactivation of the catalyst in bed 4
seems unlikely to have caused the excursion for four reasons: (1)the bed 4 excursion happened very
quickly,(2)the bed 4 temperature rose higher than a restoration of catalyst activity would have caused;(3)
no other beds(which were similarly affected by nitrogen)showed an increase*in activity,and(4)the
excursion was highly localized. Uneven catalyst activity resulting from the original catalyst loading was
ruled out for similar reasons. Loading of a comparatively high activity catalyst,using the dense loading
method,would have resulted in a whole layer of unusually active catalyst. This would have resulted in a
more uniform high temperature excursion,occurring throughout the bed, rather than the localized hot
spot evident during the incident.
(33)
Even with good liquid distribution, it is known that localized hot spots can evolve
spontaneously in the course of highly exothermic reactions, as a result of physical property
variation with temperature and composition. Thus, one source of the temperature
instability may have been the spontaneous evolution of a hot spot.
Another possibility is that the flow regime in bed 4 contributed to formation of a
"hot spot." Reactor systems employing fixed catalyst beds with downward concurrent
flow of liquid and gas reactants are known in the chemical process industry as"trickle bed
reactors." Research on trickle bed systems indicates that liquid distribution to the inlet of
the catalyst bed is critical to temperature stability. Beds in which some of the catalyst is
wet and some of the catalyst is dry are particularly susceptible to hot spots since the
reaction rate and mechanism of heat generation and removal are different for wet, partially
wet and dry catalyst. Temperature gradients are more sensitive to liquid distribution in
this transition zone than in flow regimes where the catalyst is either completely wet or dry.
Flash calculations indicate that the reactants transitioned from wet to dry catalyst within
the 4th bed of Reactor 3.
Finally, it appears from the 'presence of the agglomerated, high coke content
catalyst pillars in the 4th bed that uneven liquid distribution had occurred at some point in
time. Though temperature gradients for the 4th bed have been generally similar to those
for the other beds, coke deposits and temperature instability in the 4th beds of the Stage 2
reactors have been noted on prior occasions. Post incident inspection of Reactor 3 also
(34)
indicated distortion of the manways and the support steel for the 4th bed distributor.
However, it was not possible to determine whether either of these conditions existed prior
to the incident, since they could have been caused by the rapid pressure decrease during
the incident. If either did exist prior to the incident, it could have resulted in uneven
distribution of liquid to the 4th bed,
The Response to the Excursion
The initial temperature alarm sounded at 7:34 pm. The explosion occurred at
7:41:30 pm, 7-1/2 minutes later. Analysis indicates that, after 7:34 pm, there were no
operating changes, short of a rapid depressuring of the unit, that would have impacted the
course of this event.
The operators did not depressure the unit primarily because they did not believe
the temperature data displayed by the data logger, and they had very little time in which to
verify that data before the pipe rupture occurred. The operators did not believe the
temperature data for a variety of reasons:
• They had prior experiences with instrument unreliability,
• They were confused by the data logger fluctuations between high temperatures and 0;
(35)
• The excursion did not conform to their expectations about how such an event should
behave; specifically, they did not expect to see high temperature increases from one
bed to the next, did not expect to see temperatures increase as rapidly as they did, and
did not expect to see reduced hydrogen makeup flow and high hydrogen purity, as
indicated by the analyzer, during a temperature excursion;
• The operators did not observe the high temperatures on the bed 5 outlet or the reactor
outlet, as indicated on the control board strip charts shortly before the explosion, in
part because they were distracted by the need to monitor trim furnace firing;
• Consistent with prior experience, they believed they had time to verify the temperature
data and control the process, rather than depressuring the unit as.required by the
procedures.
Recommendations:
• Take steps to provide more uniform distribution to the catalyst beds.
(36)
• Provide for automated activation of the 300 psi depressuring system from the unit
control system based on bed thermocouples and from an independent system based on
reactor outlet temperatures.
• Replace the data logger with a DCS based system connected to all 96 reactor bed
thermocouples.
• Install high priority, highly visible temperature alarms on the reactor effluent line.
• Improve the response time of the hydrogen analyzer.
• Display the amperage of the recycle compressor on a primary Stage 2 process display.
• Review, re-issue and retrain on operating procedures related to temperature
excursions at the hydrocracker.
• Address job performance issues with the involved operators, supervisors, and technical
personnel.
• Re-train personnel to understand that undetected hot spots can appear in an upper bed
leading to an apparent rise in temperature between beds.
(37)
B
• Re-train personnel on unit instrumentation.
• Re-train plant personnel on the behavior of the hydrogen system in the hydrocracker.
• Identify high priority critical operating limits and review operating data frequently to
determine if plant operating conditions are outside of these limits. Investigate these
events and consistently correct equipment, procedural and job performance issues that
relate to these events.
(38)
ATTACHMENT A
72-Hour Written Follow-up Notification
Hydrocracker Explosion and Fire
Tosco Avon Refinery
January 21, 1997
This report on the January 21, 1997 Hydrocracker explosion and fire at the Avon Refinery is
submitted to the Contra Costa County Health Services Department at its request.
The Incident
On January 21, 1997, at approximately 1944 hours, an explosion and fire occurred within the
Hydrocracker Unit at Tosco's Avon Refinery. This incident resulted in the death of a Tosco
employee, an Operator who was on duty at the Hydrocracker. It also resulted in injuries to an
additional 25 people, including seven Tosco employees and 18 contractor personnel working at
Tosco who were on site at the refinery at the time of the incident. These injuries included a
fractured foot, emotional trauma, and more minor physical injuries such as headaches, ringing
ears, minor cuts and scrapes, and twisted knees. Thirteen personnel were transported by
ambulance or other means to various area hospitals, including Mt. Diablo, County, John Muir
and Kaiser. Others went home or returned to work. All of the individuals transported to
hospitals for treatment were subsequently released; there were no hospital admittances required.
There were no reported injuries to the public or any offsite personnel as a result of this incident.
The wind during the initial stages of the incident was out of the south-southwest. As a result,
most of the smoke resulting from the fire was carried over the uninhabited area north of the
refinery and Suisun Bay. It does not appear that there was any significant community impact
resulting from the smoke. The wind speed was estimated at 5-7 mph during the initial stages of
the incident.
Material Involved
The material involved in the explosion and fire was the combined product from one of the
Hydrocracker Stage 2 reactors consisting of butane, light.gasoline, heavy gasoline, gas oil and
hydrogen. Because the feed to the second stage is hydrotreated (desulfurized), the material does
not contain hydrogen sulfide or other acutely hazardous materials (AHM's) as defined in the
State's Risk Management and Prevention Program (RMPP).
Emergencv Notifications
Within minutes of the explosion, the Shift Superintendent called Tosco's Security Control and
requested activation of Tosco's emergency procedures and immediate notification to the Contra
Costa County Health Services Department (CCCHSD) and the California Office of Emergency
Services (CAOES).
Upon witnessing the explosion, Security Control had already begun to activate the emergency
procedures by sounding the fire horn. Immediately following the explosion, all five of
Security's phone lines were activated with incoming calls. This presented two problems. First,
due to the sheer volume of incoming calls, Security was overwhelmed. Secondly, it left no
outside lines available to make outgoing phone calls, including agency notifications. However,
Security Control was able to make the following emergency response notifications:
At approximately 1947, Security Control activated the numeric Tosco first responder
pager.
At approximately 1947, Security Control called 911 and requested four ambulances to
stand by.
At approximately 1952, Security Control activated the alphanumeric Tosco first
responder pager.
The Tosco Emergency Command Center (TFCC) was also activated within minutes of the
explosion. From TFCC, additional notifications were made independently of those made from
Security Control. At approximately 2015, a Tosco representative called the CCCHSD
emergency reporting phone number (646-1112). The call was answered by a Sheriff's
Department emergency dispatcher. The Tosco representative informed the dispatcher that he ,
was attempting to reach CCCHSD to report an emergency at the refinery. The dispatcher
inquired about the nature of the emergency and was informed about the explosion and fire. The
dispatcher informed the Tosco representative that the phone had rolled over to him, and that both
the Sheriffs Department and CCCHSD were aware of the incident and were en route to the
refinery.
Additional subsequent agency notifications made from TECC included the following:
At approximately 2010, the Bay Area Air Quality management District was notified. A
voice mail menu was reached, the option for emergency breakdowns was selected, and a
message was left.
At approxninately 2026, Mr. Jim Hatturn of CCCHSD was informed of Tosco's request
to activate the Community Alert Network and sirens for a Level 3 incident.
At approximately 2030, the California Office of Emergency Services (CAOES) was
notified. Mr. Bill Rennington took the call (report number 97-0302).
At approximately 2040, the California Department of Fish and Game was notified. Lt.
Bono took the call.
At approximately 2050, the National Response Center (NRC) was notified. Mr. Bright
took the call (report number not available).
At approximately 2103, EPA Region IX was informed (Mr. Dunkelman).
At approximately 2235, Cal-OSHA was informed (Mr. John Cayiak).
The Santa Fe railroad was also notified of this incident. See the attached summary for further
information regarding emergency notifications.
Emergency Response
Tosco activated its employee volunteer fire brigade for this incident. The Incident Commander
was on site at the time of the explosion and immediately set up an Incident Command post.
Personnel were sent to the scene to assess the situation, and to the Hydrocracker control room to
check on the Operators. At that time, they learned that the unit was being shut down, isolated
and depressured, and that one Operator was missing. Tosco also began a systematic search for
all employees and contractor personnel throughout the refinery.
Response consisted of applying cooling water to the affected area and surrounding structures
using firefly portable fire monitors and all of the stationary fire monitors in the area. Additional
fire pumps were started throughout the refinery as required to maintain fire water pressure. The
fire was contained to the Hydrocracker Stage 2 No. 3 Reactor outlet, control valves and
associated piping, and allowed to bum out. Approximately 50 Tosco firefighters were utilized to
contain and control the fire.
Incident Investigation
The explosion appears at this point to have originated in the outlet piping from the Hydrocracker
Stage 2 No. 3 Reactor, at a point just prior to entering the feed-effluent heat exchangers. The
root cause of the incident is unknown at this time. Tosco is conducting an investigation of this
incident, and is cooperating with investigations being conducted by several agencies. Tosco will
keep the County informed of the progress of its investigation.
Summary of TFCC Notifications
Hydrocracker Explosion and Fire
Tosco Avon Refinery
January 21, 1997
Approximate
Time
1944 Explosion and fire at Hydrocracker.
1945 Security Control sounds fire horn.
1947 Security Control activates the numeric Tosco first responder pager.
1947 Security Control calls 911 and requests four ambulances to stand by.
1947 Shift Superintendent calls Security Control and requests notification of CCCHSD
and CAOES.
1952 Security Control activates the alphanumeric Tosco first responder pager.
2010 TECC notifies Bay Area Air Quality Management District.
2015 TECC notifies CCCHSD at phone number 646-1112. Phone rolls over to
Sheriffs Dispatch, who informs Tosco that the Sheriff's Department and
CCCHSD are aware of the incident and are en route to the refinery.
2020 Two ambulances standing by.
2026 CCCHSD (Jim Hattum) calls. TECC requests that CCCHSD activate the
Community.Alert Network and sirens for a Level 3 incident.
2030 TECC notifies the California.Office of Emergency Services. Report number 97-
0302.
2037 Tosco receives first Community Alert Network notification.
2040 TECC notifies California Department of Fish and Game.
2041 Tosco receives second Community Alert Network notification.
2047 Sheriff and Consolidated Fire on stand-by at D Street Gate.
2050 TECC notifies the National Response Center.
Approximate
Time
2103 TECC notifies EPA Region IX(Mr. Dunkelman).
2136 Faxed product MSDS to local hospitals.
2140 Returned call to Linda Pappeson of California State Fire Marshal's office.
Reported that no jurisdictional pipelines were involved.
2209 Cal-OSHA(Mr. John Cannec) calls and is updated on incident by TECC.
2245 Received call from CAOES (Jim Maquis). Updated on incident. Asked for call
back if something significant changes.
2248 Fire Chief reports deceased located.
2300 On-site representatives from CCCHSD (Charles Nicholson) and BAAQMD
(Mike Weedl) notified of fatality.
2300 Coroner notified.
2315 Coroner on site.
2333 Cal-OSHA(Rick Ullerich) notified of fatality.
2347 Department of Toxic Substance Control (Allan Freihoffer) offered assistance.
0055 All Clear.
E F —
^
Zr N �D Aye
a
9
�d A
�a ^
�a N
1
1 1 a R
a.w5 1
N �r w
:J...
/I/ I� `� I II /t\ I/ \ 1\I
� 1 I
a ^ �
65s,f (
152'f
m P �
wf ss`rr
n w N w
� 1
g � 1
CI
PC.
N� N
i t N
HSA Nidi' Cl
� O
Q eR
G7 NH m
1"-
I^� vM
0 uH ^
r.
gE0
8ED
taG
tao
BED
taj
BED
5 �
e2�
SCO REFINING COMPANY't
a,RM14 v VARM4 ,a
MC niERUOCOUPIE ARRANGEMENT
STAGE 2 REACTORS
STAGE WDROCRACKER(AWO)"IrA u. a sr Mer r
a FIGURE 2
1 BOARD OF SUPERVISORS
2 CONTRA COSTA COUNTY C(D
MCD
3 BOARD CHAMBERS
4 MARTINEZ, CALIFORNIA
5 - -000- -
6
7 IN RE : ITEM D . 7 - CONSIDER REPORTS FROM )
HEALTH SERVICES DIRECTOR AND FROM )
8 TOSCO AVON REFINERY ON FINDINGS )
OF INVESTIGATION OF JANUARY 21 , 1997 )
9 HYDROCRACKER INCIDENT )
10
11
12
its 13
TRANSCRIPT OF TAPE-RECORDED PROCEEDINGS
14
MEETING OF . JUNE 3 , 1997
15
16
17 PRESENT :
MARK DeSAULNIER, Chairman
18 JIM ROGERS , Supervisor
GAYLE UILKEMA, Supervisor
19 DONNA GERBER, Supervisor
JOSEPH CANCIAMILLA, Supervisor
20
21
22
23
24
25
Certified Shorthand Reporters
'JAMnikolmna 1 2321 Stanwell Drive•Concord.CA 94520-4808
! REPORTING SERVICE.I1VC. P.O.Box 410;•Concord.CA 94524-4107
(510)685-62,2•Fax(510)685-3829
1 PROCEEDINGS JUNE 3 , 1997
3 CHAIRMAN DeSAULNIER : That will move us to D . 7 ,
4 which is to consider reports from the Health Services
5 Director and from Tosco Avon Refinery on the findings of
6 their investigations. of the January 21 , 1997
7 hydrocracker incident .
8 And we ' ll begin with Dr . Walker .
9 DR. ,WALKER : Mr . Chairman, the presenter would
10 like to set up an easel , if we could take about a
11 two-minute stretch for him to do that . Then we won' t
12 disrupt the conversation .
• 13 CHAIRMAN DeSAULNIER : Okay. Two minutes ..
14 DR . WALKER : Okay'.
15 CHAIRMAN DeSAULNIER : And for those members of
16 the public who would like to speak on this item., there
17 are yellow speaker requests either up here in front of
18 the chambers or in the back . You need to fill those out
19 and submit them, please .
20 So two minutes .
21 (Short break taken . )
22 All right . We will begin . Dr . Walker .
23 DR . WALKER : Mr . Chairman, members of the Board,
24 I ' ve been before you a couple of times in the last
25 couple of months with regard to issues that staff from
JfZal tdormI B Certified Shorthand Reporters
LLK�ll l` 2 232] Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. I P.O.Boa 4107•Concord.CA 9452.1-4107
(510)685-6222•Fax(510)685-3829
1 the Hazardous Materials Division of the Health
• 2 Department were having with. regard to access to
3 information from Tosco regarding the January 21st
4 hydrocracker incident .
5 I ' m pleased to inform you today that your efforts
6 and your support have been successful in our -gaining
7 access to the information and the interviews that we
8 needed .
9 And we are presenting to you two reports today,
10 one from, Tosco and one from our Health Department staff
11 with regard to the root cause analysis of the
12 January 21st incident . Both of these reports have been
13 released to the press and to the public . We will go
• 14 over those in
sufficient amount of detail today, and
15 I ' ll have some closing comments after that .
16 But I ' d like to first begin with introducing
17 Dwight Wiggins , the CEO of Tosco, and John Miller, the
18 refiner manager, who will present Tosco' s report .
19 CHAIRMAN DeSAULNIER: Welcome .
20 MR . WIGGINS : Good afternoon . I appreciate the
21 opportunity to address the County Board of Supervisors .
22 I am Dwight Wiggins . I ' m president of Tosco Refining
23 Company.
24 And in order to provide you with a very brief ,
25 brief background of myself , I ' m a graduate engineer . I
JZandoneuc Certified Shorthand 520-4 08
L�1 ll 1 3 2321 Stam��ell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC; ; P.O.Bos 4107•Concord,CA 94524-4107
(510)685-6222• Fax(510) 685-3829
1 have over 30 years of experience in the petroleum
• 2 refining industry, and I ' ve been a line manager in the
3 refining industry for over 15 years , including seven as
4 refinery manager of a Bay Area refinery: And I lived in
5 Supervisor Gerber' s district .
6 Throughout my entire career, I believe that three
7 of the most important objectives of operating a refinery
8 were to provide a safe workplace for the employees and
9 the surrounding community, to prevent fires , and to
10 prevent explosions . On the night of January 21st , 1997 ,
11 unfortunately this was not to be .
12 Today I ' m speaking not only for myself , but for
13 all the men and women at Tosco when I say that we ' re
i
14 extremely sorry for the accident that occurred on the
15 January 21st . It ' s deeply affected each one of us , and
16 we lost a respected coworker, Michael Glanzman . Tosco
17 acknowledges and accepts responsibility for providing a
18 safe workplace.
19 The accident investigating period has been very
20 difficult for all of us . I was in Phoenix, Arizona at a
21 business meeting on the night of January 21st when the
22 accident occurred. As soon as I was informed that an.
23 individual was - missing, I made immediately plans to fly
24 to Avon and I arrived early on the morning of January
25 the 22nd .
,,,� Certified Shorthand Reporters
Mndor Ha 4 2321 Stanwell Drize•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 410 0 Concord,CA 9.1524-4107
1 (510)685-6222•Fax(510)685-3829
1 My purpose in arriving was to first of all make
• 2 sure that the injured people and the family of Michael
3 Glanzman were being cared for to the extent it was
4 ' possible , and to make sure that a thorough investigation
5 began immediately to determine what happened, why it
6 happened, and how to use every lesson learned to make
7 sure that it does not happen again .
8 I ' m confident as I stand here today that we do
9 understand what happened and how to prevent it from
10 happening again at Tosco . And, in addition, I ' m
11 committed to sharing what we have learned and so that we
12 can educate others in the industry to prevent it -from
13 happening anywhere else .
t 14 I have great confidence in our refiner manager,
9 Y g ,
15 John Miller, and in his absolute commitment to safe and
16 responsible operations . He ' s proven to us .that he ' s
17 capable of doing this in the past four years' as refinery
18 manager of our Ferndale , Washington, refinery and during
19 his prior 20 years at Avon, that he knows how to operate
20 safely and responsibly. And I believe that he can prove
21 it to you, too .
-22 John and his refinery team have my full support
23 and direct access and that of the entire Tosco
24 corporation .
25 While John was not back home at the time of the
Certified Shorthand Reporters
�MndUI dUJ��
55 2321 Stanwell Drize•Concord.CA 94520-4808
REPORTING SERVICE. I P.O.Bos 4107•Concord.CA 94524-4107
1 (510)685-6222•Fax(510) 685-3829
1 incident , he' s been here since March and has closely
• 2 followed the investigation . He ' s the one charged with
3 implementing the recommendations and moving us forward .
J
4 from here .
5 . I ' ve asked him to share with you details of the
6 investigation and how we intend to use the information
7 that we ' ve collected. Before I ask John to come
8 forward, though, I ' d like to once again assure you that
9 Tosco is totally committed to safety, we ' re committed to
10 strengthening the relationships with the County, with
11 other agencies and with the community at large . And we
12 are committed to restoring the public trust and
13 confidence in Tosco .
• 14 I thank you for this opportunity to speak .
And
15 at this time I ' d like to ask John Miller to come
16 forward . And then we ' ll be available for whatever
17 questions you have .
18 Thank you .
19 CHAIRMAN DeSAULNIER : Thank you, Mr . Wiggins .
20 Mr . Miller .
21 MR . MILLER : I ' d like to thank the Supervisors
22 for this opportunity to come and talk to you today about-
23 the results on our investigation report .
24 Briefly let me introduce myself . I ' m a native of
25 the area , born and raised in California, an engineer by
��r^��O�g■ia Certified Shorthand Reporters
LKl l■ 6 2321 Stanwell Drive•Concord.CA 94520- 805
REPORTING SERVICE.INC. P.O.Hoy 410;•Concord,CA 94524-4 107
i (510)655-6__:.222.Fax (510)685-3829
1 original training. I have over 20 years ' experience in
2 the refining business . The -most recent , as Dwight
3 explained, was Ferndale refinery manager in the State of
4 Washington .
5 And as of March 1st of this year, I assumed the
6 role of plant manager for the Tosco Avon refinery. And
7 with the completion of the acquisition of 76 Products
8 Company, I became the general manager for the refineries
9 at both Rodeo and Avon, which we call the San Francisco
10 area refinery.
11 We all at. Tosco regret the very tragic incident
12 that occurred January 21st of this year . We sincerely
13 apologize to all those who were affected by this
• 14 incident .
15 Believe me , no one cares more than the men and
16 women of Tosco on finding out what happened, why it
17 happened and, more important , what are we going to do to
18 prevent it from ever happening .again .
19 As Tosco' s vice president of refining and the
20 general manager for the San Francisco area refinery, I ' m
21 responsible to see that both facilities are operated and
22 maintained safely. I take- this responsibility very
23 seriously, and I expect my management and. my workers to
24 do the same . When unsafe conditions or accidents occur,
25 it ' s my job to ensure that actions are taken to
Certified Shorthand Reporters
�ZaradoraeIIa 7 2321 Stanwell Drive•Concord.CA 94520-4808
REPORnNG SERVICE.INC. P.O.Box 4107•Concord.CA 94524-4107
(510)685-622.2•Fax (510)68.5-3829
1understand what happened and to solve the problem once
2 and for all .
3 . .We recently completed our report on a fairly
4 extensive and thorough study which culminated almost
5 four months from start to beginning (sic) . And this
6 report has been distributed, and I hope you' ve had the
7 opportunity to review this report .
8 Today I ' d like to go over this report and discuss
9 some of the more important highlights . I ' d like to
10 begin .with reviewing basically what is a hydrocracker,
11 what it does , to try to frame everybody' s mind on the
12 kind of equipment and process we ' re talking about .
13 Then I ' d like to discuss about the - - how the
14 incident itself was organized and carried out . And then
15 I ' d like to discuss the corrective actions that are
16 planned and actually will be implemented or in progress
17 as we speak .
18 We ' ve taken the liberty to bring a couple
19 drawings of the unit to try to illustrate the unit that
20 we ' re talking about . And I ' have Chuck Waitman here who
21 will be assisting me . And he actually is our safety
22 manager who headed up the investigation .
23 The unit description, let ' s start out with the
24 hydrocracking process itself . This is one of many
25 processes in the refinery for converting crude oil to
ii Certified Shorthand Reporters
I
�Zandon 18 - , 8 2321 Stanwell briJe•Concord,CA 94520-4808
-
REPORTING SERVICE.INC.. P.O.Box 107•Concord,CA 93524-4107
(510)685-6222•Fax(510) 685-3829
1 clean products . In the case of a hydrocracker, it
• 2 processes a heavy oil stream, which is something between
3 kerosene and lubricating oil . And it -actually converts
4 it to diesel and gasoline products .
5 There' s many different types of design. Ours is
6 one . The drawing on the right , as you can see, there
7 are six large vertical vessels . There ' s actually two
8 stages in our design .
9 The first three vessels on the right are all in
10 parallel . That ' s the first stage of the unit , which
11 actually serves to prepare the oil for the second stage .
12 Its primary purpose is to remove contaminants such as
13 sulfur and nitrogen which deactivate the catalyst which
14 is in the second unit .
15 we will be talking a little bit about. catalyst
16 and catalyst beds today, and I actually brought a sample
17 to try to illustrate - - this is available for
18 inspection - - at least one aspect of what is actually in
19 the unit .
20 The second stage of the unit which is made up of
21 another three parallel reactors , these reactors are
22 actually fairly large . They' re diameters are 78 feet
23. and they' re over 100 feet tall . And the second stage is
24 the cracking part of the unit . This is where the heavy
25 oils are actually broken down into lighter oils ,- which
�M,,� Certified Shorthand Reporters
udoi MA 9 2321 Stanwell Drive•Concord,CA 94520-4808
REPoiznliG SERVICE.INC. P.O.Box 410"•Concord.CA 94524-4107
(310)685-6222• Fax (510)685-3829
1 in the case of this unit is designed to make gasoline
2 and diesel products .
3 The drawing . to the left is just an outline.
4 drawing of the reactor that was actually involved in. the
5 incident . It ' s a second stage cracking reactor . There
6 are five catalyst beds in this reactor which are labeled
7 on the drawing .
8 We feed a combination of oil and hydrogen to this
9 unit which are mixed and preheated in the heat
10 exchangers that are shown on the drawing . And then
11 what ' s not shown on the drawing is there is actually a
12 heater loop that takes a slipstream into hydrogen . And
13 this oil enters the top of the reactor . It ' s a
14 down-flow reactor so therocess proceeds from the to
P P P
15 to the bottom of the reactor .
16 At the top of the reactor, at the top of each
17 bed, there are distributor trays . The oil as it enters
18 with the. hydrogen is actually in two phase , meaning that
19 partially it ' s gaseous and the other part is liquid.
20. The distributor trays are designed to evenly distribute
21 the gas and liquid across the bed of the reactor .
22 And in the catalyst bed is where the actual
23 cracking reaction takes place . This reaction is what ' s
24 commonly called exothermic . That means it generates
25 heat . So you actually get a temperature rise as you go
i Certified Shorthand Reporters
�ZandOROUB 'n 10 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. j Y.U.Box 410.•Concord.CA 94524-410
(510) 685-6112 0 Fax (510) 685-3829
1 from the top to the bottom of each bed.
• .2 As you notice , off to the right we have several
3 lines that are called quench hydrogen . Quench hydrogen
4 is brought in to both replace the hydrogen that is
5 consumed in the reactor as the oil is cracked. They .
6 absorb hydrogen. And it ' s also to cool the reactants
7 leaving the bottom of each catalyst bed to bring the
8 temperature back to a temperature that ' s optimal for the
9 operation of this unit .
10 And then, of course, at the bottom of the
11 reactor, the final products , which are basically diesel
12 and gasoline , leave the reactor . That ' s the hottest
13 point of the reactor normally, and they enter heat
• 14 exchangers which exchanges heat with the
g g colder incoming
15 feed .
16 The top of one of those lines is where the actual
17 failure occurred in this unit , which Chuck is pointing
18 to on the , what ' s referred to as the fluent line inlet
19 to the feed and fluent heat exchangers . That ' s a brief
20 overview of the process itself .
21 A little bit about the history of the unit . The
22 unit was originally designed and built in 1963 . It was
23 built to a Chevron license called an isocracker, which
24 was a two stage design as I described .
25 The unit over its history has been gradually
ZQRikormna Certified Shorthand Reporters
VQ ■l 11 2.321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4107•Concord,CA 94524-4107
- (510)685-6222 0 Fax(510)685-3829
1 upgraded as the technology evolved . In . 1986 , _ the
• 2 process was converted to the state-of-the-art technology
3 which is a unicracking process . This was licensed by
4 Unocal and presently licensed by UOP.
5 These units go through maintenance cycles of
6 roughly two to four years . In other words , every two to
7 four years , these units are shut down to do what we
8. refer to as a turnaround which is major maintenance .
9 And what this major maintenance means is you
10 basically overhaul the equipment to make it very much
11 like new so that you can operate your next two to
12 four-year cycle safely and reliably .
• 13 The key to operating these units in all
14 refineries . is you continuously upgrade the units as
15 technology evolves .
16 Also- part of making sure that your units are
17 operated safely and reliably during its run between
18 turnarounds is that you do extensive on-line and
19 off-line inspections . And that ' s to make sure that any
20 aging or corrosion of equipment is identified and
21 corrected at the proper time to maintain safety of the
22 unit .
23 As far as staffing on this unit goes , until
24 somewhere between five and ten years ago, I don' t
25 know - - remember the exact date , this unit was staffed
� Certified Shorthand Reporters ZaradoraeIIa 12 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Roz 410 •Concord.CA 94524-4107
(510)683-6222• Fax(510) 685-3829
1 by four operators . Like I say, a little over five years
• 2 ago, the decision was made to increase the staffing on
3 this unit to five operators . So rather than having all
4 operators .with dedicated areas of responsibility, we
5 created a head operator position that had no specific
6 equipment responsibility in the unit , but was
7 responsible for coordinating overall activities .
8 To qualify as an operator on this unit requires
9 an average of 18 months of training . This training
10 involves everything from textbook to what we refer to as
11 job shadowing, which is a one-on-one mentoring out in
12 the field on the actual equipment , and a large number of
13 other ways of presenting the information to the
® 14 operators .
15 And of course , before they become qualified, they
-16 are tested, both written, verbally and demonstrated in
17 the field .
1.8 This unit is a major producer of Garb gasoline
19 and diesel . It produces components that ultimately
20 become it . And it is one of the units that ' s the heart
21 of our refinery.
22 This unit has been shut down since the explosion
23 and fire in January and is in the process of being
24 rebuilt , incorporating many of the improvements that
25 we ' ll be discussing later . And right now we tentatively
Certified Shorthand Reporters
�Za ndonella 1 13 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.B6x 4107•Concord.CA 94524-4107
(510)685-6222•Fax (510)685-3829
i
1 look to restart this unit by about July of this year .
• 2 That ' s basically an overview on the hydrocracker
3 process . I ' d like to discuss briefly the
4 instrumentation on this unit .
5 Referring back to the drawing on the left , the
6 reactor, is that there are a large number of .controls on
7 this unit . What ' s not shown on the diagram, but all the
8 feed streams , there ' s flow controls , and on all the
9 catalyst beds., both inlet and outlet , there are an
10 enormous number of temperature monitoring devices which
11 are normally called thermocouples , or we refer to them
12 commonly as TI ' s .
13 So at the inlet and the outlet of each bed, there
14 is a TI . And this instrument is critical for operating
15 this unit safely and efficiently.
16 What the operators actually have their hands on
17 are instruments that are located both in a control room,
18 which is the control center for this unit , which is
19 located on the plot space of the unit , and local
20 instrumentation .
21 These instruments are a combination of what are
22 referred to as board mounted, in other words these are
23 instruments that are located actually on a wall in the
24 control room which control certain control loops such as
25 the temperature to the inlet at the top of any one of
Certified Shorthand Reporters
�ZandOEMfl8 14 2321 Stanwell Drize•Concord.CA 94520-4808
i REPORTING SERVICE.INC. P.O.Box 410;•Concord,c:A 9452 -410;
(510)685-62D2 0 Fax(510)685-3829
1 the beds or the temp- - - or the flow to . any one of the
. 2 hydrogen quenches , in combination with a couple of
3 levels of computer control , which there ' s a data logger
4 that controls a large number of temperatures throughout
5 the catalyst beds and the reactor .
6 And there is also an IA that for the second stage
7 was designed to do essentially the same purpose . An IA
8 is a fairly sophisticated distributor control system,
9 which is an abbreviation for intelligent automation .
10 That ' s a little bit about the design .
11 The history of the instrumentation, when this
12 unit was built in 163 , it actually had one of the first
13 electronic control systems at the time . It was a very
• 14ro ressive design . At the time most of the units w
P g g were
15 controlled by pneumatic controllers .
16 Continuous with upgrade of the equipment over the
1.7 years , the instrumentation is also upgraded as the
18 technology progresses . And we ' ve made a number of
19 improvements over the years . Most recent was the
20 installation of the board mounted controllers which are
21 called Moore controls . These effectively are little
22 mini computers that can be programmed todo different
23 kinds of control functions within the unit .
24 And of course , what we did during our turnaround
25 in 1996 , which I believe took place in January - - is
r Certified Shorthand Reporters
ZandoneIIa -1 15 2321 Stanwell Drive•Concord.CA 94520-4808
1 REPORTING SERVICE.INC. P.O.Box 410;•Concord.CA 94524-1107
-- -- -.-� (510)685-6222•Fax (510)685-3829
1 that correct , Chuck - - is that by further improvement ,
• 2 we actually by our own discretion, decision to improve
3 monitoring the unit and control of the unit ,. increased
4 the number of thermocouples in the reactor from- 40 ,
5 which was essentially state-of-the-art originally, to
6 96 , which we decided that this improvement was important
7 to better understand and control our process .
8 40 of those temperature points were brought into
9 the control room as existed prior to the turnaround in
10 January 1996 . And the remaining 56 temperature points
11 were located in a local instrument panel which was in
12 the field adjacent to the reactor .
13 The basic reason for this was , is that when,you .
® 14 do upgrade , you have a tendency, particularly when
15 you' re dealing with sophisticated control systems , to
16 phase your project . So in various stages , all of the
17 temperature points were going to be brought into the
18 ' control room at some point .
19 And, of course, as far as instruments that we
20 continue to use , is that there was a data logger that is
21 in place that monitors many of the temperature points in
22 the reactor . This is located in the control . room. And
23 on that system, that ' s where our alarms are that alarm
24 the operators when a temperature falls outside its
25 normal- operating range .
Certified Shorthand Reporters
�_ mRdUItd 16 2321 Stamkell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Bos 4107•Concord.CA 94524-4107
(510)685-6222•Fax (.510)685-3829
1 That ' s basically an overview on the unit
• 2 description . Now. I ' d like to talk about the
3 investigation itself .
4- I ' d like to start out with first the process .
5 And of course, the beginning of the process really is
6 how is the investigation structured .
7 Immediately after the explosion and the fire on
8 the 21st , an investigation team was put together . This
9 was basically our refinery and corporate policy on'
10 investigating all incidences that cause or could cause
. 11 loss . And it ' s also part of our process safety
12 management program, commonly known as PSM.
13 That team was headed up by Mr . Waitman here .
14 It ' s a team which is a multi-functional team. It
15 includes a variety of technical resources . It
16 Includes metallurgical experts , includes safety experts .
17 It also includes an OCAW health and safety
18 representative .
19 And this is a broad range team that we bring to
20 bear on any kind of incident to make sure that we get in
21 all different perspectives and can . thoroughly
22 investigate any kind of incident , particularly the more
23 complex ones which this one proved to be .
24 In addition, because of the magnitude of the
25 incident , there were a large number of agencies that
qzQ^ ]� Certified Shorthand Reporters
niloneIIB 17 2321 Stanwell Drive•Concord,CA 94520-4808
(510)685-6222• Fax(510)685-3829
REPORTING SERVICE.INC: P.O.Box 410,7•Concord.CA 94524-410;
1 came in and ultimately became part of the investigation .
2 And, of course , these agencies included both local ,
3 - state and federal .
4 Locally we had. the Department of Health Services ,
5 Bay Area Air Quality Management District , Cal OSHA, and
6 we had EPA. and Fed OSHA. These agencies got together
7 and worked out a protocol and also established a lead
8 agency. And because of the nature of the incident , Cal
9 OSHA was selected, or selected themselves , . whatever the
10 case may be , as the lead agency.
11 An investigation protocol was established whereby
12 information would be made available to all interested
13- parties and there would be some control over how the
• 14 information wasathered . And this is absolutely
g to y
15 critical in an investigation, is to control the quality
16 of the information .
17 This information was shared with all parties
18 concerned, all the agencies . And, of course., Tosco was
19 the key player in supplying large volumes of information
20 to all the agencies . This included not only the
21 history, the design, the inspection records , the
22 maintenance records , the process data, everything up to
23 and including the interviews themselves .
24 _ That ' s a little bit about the structure of the
25 investigation . Now some of the details and focus that
Certified Shorthand Reporters
�ZandUItPsTla 18 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTTriG SERVICE.INC. P.O.Box 4107•Concord,CA 94524-4107
i'
(510)685-622°_0 Fax(510)685-3829
I the investigation keyed on .
2 We gathered all the pertinent operating data
•
3 before and during the incident . There' s a large number
4 of sources , both on various computer data bases , manual
5 strip charts , logs and other sources . There are
6, literally hundreds of these sources .
7 Also , the operators involved, both directly and
8 indirectly, and supervisors involved, both directly and
9 indirectly of this unit , were all interviewed as to the
10 incident and factors leading up to the incident .
11. The maintenance records , inspection records were
12 inspected in very large amount of detail , being a very
13 critical part of the investigation .
• 14 And also after the incident , we thoroughly tested
g Y
is all of the instrumentation involved around this reactor .
16 This included the temperature points and the various
17 recorders and indicators in the control room.
18 We thoroughly did a literature search and
19 literally an industry search of all the latest
20 information on hydrocracker technology, with particular
21 focus towards the kind of event that occurred at our
22 unit. A very key part was that all the. equipment was
23 thoroughly inspected after the incident , as access could
24 be made to not only the pipe that failed, but the
25 reactor itself , internals , worked very close with Cal
ZartdoraeIIaCertified Shorthand Reporters
19 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC P.O.Box 1107•Concord.CA 91524-4107
(510)655-6222• Fax (510)655-3829
1 OSHA on this throughout to make sure all that
• 2 information was gathered to the best of everyone' s
3 ability .
4 And a lot of these materials were sent to outside
5 labs for thorough analysis as far as the pipe that
6 actually failed., its metallurgical analysis and its
7 condition relative to hardness , softness , and all the
8 other factors that the metallurgists look to do failure
9 analysis on that pipe , in addition to a lot of analysis
10 on the catalyst itself .
11 Part of this is that we did a very thorough
12 analysis of the data , itself , and part of it was , is to
13 do a lot of modeling or very say thorough and complex
• 14 calculations to try to reproduce empirically what we
15 found in the field to make sure that , from an
16 engineering and all other standpoints , that we could
17 verify the confidence level of the information gathered .
18 Also review of the operator training material and
19 qualifications , - and probably equally important is we did
20 a thorough survey of the industry. With the acquisition
21 of 76 Products Company, this added two more
22 hydrocrackers to Tosco and we were able to further bring
23 this information into our investigation as far as the
24 lessons learned in these units and how these units were
25 operated and maintained .
z�. i Certified Shorthand Reporters
andOIMIlB C. 20 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.IN
' P.O.Box 410 •Concord.CA 94524-4107
`_— - - --- -.---- (510)685-6222•Fax(510)685-3529
1 The culmination of these activities took a little
• 2 over three months to both gather and analyze this data .
3 That was a little bit about the process . Now I ' d
4 like. to talk a little bit about 'some of the findings .
5 To kind of frame things , I ' d like to start out with
6 chronology of events , both prior and during the actual
7 incident .
8 The unit status. As I mentioned earlier, this
9 unit is normally staffed by five operators . The average
10 experience of the operators is six years . And of the
11 operators that were on shift that night , at least a
12 couple of them had over ten years ' experience on this .
13 unit .
• 14 In addition we had two extra operators on . The
15 reason for that is that there were .a number of
16 additional activities going on .
17 A couple days prior to the incident there was a
18 leak on a heat exchanger which necessitated pulling a
19 feed from the A reactor, which is actually the far end
20 of all the reactors - - far side to the right , Chuck .
21 It ' s at the opposite end from the Number 3 reactor where
22 the incident occurred . And this was done because the
23 leak created a condition that it was necessary to pull
24 feed to operate the unit safely.
25 In order to repair this leak - - and just to share
Certified Shorthand Reporters
1 MRdoraella 21 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4'1-07•Concord.CA 94524-4107
N (510)68.7622 • Fax (510)685-3829
1 with you the history, is that there was an engineered
• 2 containment device that was installed on feed and fluent
3 exchangers of A reactor* - - Chuck, i,f you _ - . I don' t know
4 if people can see it . But it ' s the flange in between
5 the two bundles .
6 And this device was specially engineered to
7 handle a small gasket leak that developed on the run .
8 It ' s very common practice in the refining industry,
9 there' s expert companies that do this . And this device
10 is injected with a sealant material which seals the leak
11 and prevents it from causing problems .
12 Well , for whatever reason, it began to leak and
13 necessitated pulling feed from that reactor . And we had
14 reinjected the sealant into this reactor . And we had to
15 allow time for it to cure, so we had no oil in A reactor
16 at the time .
17 And we had basically two operators brought in
18 extra to heat .up this reactor, make sure that the clamp
19 didn' t leak, that it cured properly before feed would .be
20 introduced at a later date .
21 Another issue caused by this. leak was that it
22 necessitated pulling feed from the first stage , which is
23 the first three reactors . They' re all parallel and
24 identical.
25 And this led to going off specification on one of
Certified Shorthand Reporters
JMnikoiwHa 22 2321 Stanwell Dri%e•Concord.CA 94520-4808
REPORTING SERVICE,INC. P.O.Box 4107•Concord.CA 94524-4107
(510)685-6222•Fax(510) 685-3829
1 the contaminants , nitrogen to the second stage .
• 2 Nitrogen deactivates the catalyst in the second stage .
3 And what this does is it basically reduces the unit ' s or
4 the catalyst ' s ability to convert the heavier oil to
5 lighter product , so it reduces gasoline production,
6 increases the yield diesel off the unit .
7 The operating . supervision prior. to this shift had
8 gotten together and developed a procedure, and that
9 procedure was communicated, whereby we were going to
10 gradually raise the reactor temperatures on the first
11 stage , reestablish. the normal nitrogen carryover,
12 there ' s always trace amounts , to the second stage , and
13 raise the temperatures on the second stage in order to
14 strip the. nitrogen and restore the activity and the
15 conversion in the second stage . And, of course , this
16 would be done slowly. And that was part of the reason -
17 of having the extra people on shift .
18 Also prior to the incident , this procedure was
19 proceeding very smoothly . The unit was very stable . At
20 the time the data logger which displays all the various
21 temperatures around the reactor that are not on the
22 board was in service . And this is the instrument that
23 alarms operators when temperatures are out of the normal
24 operating range . And also the unit was at a minimum
25 design rate . But case in point , the unit was stable
Certified Shorthand Reporters
�Zandorad 23 2321 Stanwell Drise•Concord,CA 94520-4808
REPORTING SERVICE.MC. P.O.Box 4107•concord:CA 94524-4107
(510)685-6222•Fax (510) 685-3829
I prior to the incident.
• 2 The incident basically occurred at 7 : 34 p .m. on
3 the 21st . This was our first indication of a problem.
4 There was an alarm in the fourth reactor cat- - - or the
5 third reactor, .fourth catalyst bed at the bottom, which
6 indicated that one of the temperatures ,that. are recorded
7 by thermocouples that penetrate into this catalyst bed
8 had gone above its limit .
9 This temperature spiked and actually returned to
10 normal over a fairly brief period of time . This alarm
11 was acknowledged . by the operators and they began to
12 respond to a developing situation .
13 About the same time , a temperature spike occurred
14 in the top of the fifth bed. In other words this
15 temperature spike occurred so quickly that normally what
16 the hydrogen quench would do would be to quench down the
17high temperature . It happened so fast that the
18 controller couldn' t react fast enough to it , and so the
19 spike first appeared in the fourth bed, disappeared,
20 then appeared in the top of the fifth bed, and then
21 disappeared again . This was something that the
22 operators had never experienced before .
23 . At this time a couple other things occurred which
24 both took the operator' s attention away from watching
25 what was really going on into this reactor and also was
q� Certified Shorthand Reporters
�Lxi����PilTB
24 2321 Stanwell Drize•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Boz 4107•Concord,CA 94524-4107
(510)685-6222•Fax(510) 685-3829
1 outside their experience, and basically anybody' s-
• 2 experience at. Tosco .
3 Apparently a reaction that nobody had ever
4 experienced or remembered experiencing began to occur
5 whereby, it ' s called demethylization started going on .
6 And this is not like the cracking reaction this unit ' is
7 normally designed to do . This is a very exothermic
8 reaction, but rather than consuming hydrogen, it
9 generates methane .
10 And what this did was- that normally .you make up
11 hydrogen to the unit as it ' s consumed. And what this
12 did was it backed out all the makeup hydrogen into the
13 unit , which the operators found quite confusing because
14 they had never experienced something like that before .
15 In addition, with some of these moves on the
16 hydrogen system, the Number 1 reactor on the second
17 stage started firing its furnace very hard. And this
18 required the operators ' attention because it had started
19 going out of its operating range, and the operators had
20 to get on to address it , get it back into control .
21 In addition, either early on or during the event ,
22 the temperature logger which records the temperatures in
23 all the catalyst beds , primarily the fifth bed, began
24 giving readings that confused the operators . They were
25 getting readings that for a variety of reasons led them
� ^ Certified Shorthand Reporters
L— �dO�e�8 X-1 25 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC P.O.,Box 4107•Concord.CA 94524-4107
(510)685-6222•Fax (510) 685-3829
1 to believe or doubt that the readings were valid .
• 2 This happened only a few minutes after the first
3 alarm occurred . At that time , Michael Glanzman went out
4 to the field as part of his normal duties to verify the
5 instrumentation in the field, to try to ascertain
6 between him and the rest of the operators what was
7 really happening to them.
8 Unfortunately, before any of the operators were
9 aware or could respond, a temperature runaway, or I
10 should say a temperature excursion occurred - in the fifth
11 bed which led to temperatures that overheated the
12 effluent pipe and ultimately led to the failure of the
13 effluent pipe at normal operating pressure . Shortly
• 14 thereafter the unit had suffered its explosion and fire
15 and was depressured as. a result .
16 That ' s a brief chronology of what happened prior
17 and during the incident . There ' s more details in the
18 report , but my discussion is going pretty long so I ' ll
19 move on .
20 Some of the facts uncovered during the
21 investigation was the analysis and inspection of the
22 pipe showed. that its mechanical integrity was well above
23 minimum. It still had relatively original wall
24 thickness , its hardness , its softness , all other
25 characteristics were deemed to be well within
Certified Shorthand Reporters
(�j 26 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. 1 P.O.Box 3107•Concord,CA 94524-4107
(510)685-6222 0 Fax(510)685-3829
1 specification for the pipe in this service . .
2 The'. control functions, we verified all the . ..
3 temperature points , the thermocouples , all 'the board
4 mounted instruments , and we found them all to be in good
5 order .
6 We were unable to duplicate the observations of
7 the operators . Don' t misunderstand me , we ' re not saying
8 that the observations of the operators weren' t what they
9 saw . It ' s just after the fact we were unable to
10 duplicate them.
11 We reviewed the staffing . The staffing level for
12 the situation in the unit was more than adequate , with
. 13 two people additional on shift . And in addition we had
• 14 very experienced crew on shift.
P
15 Early on, one of the possibilities that we
16 investigated was the nitrogen contamination of the
17 second stage . And we investigated that very thoroughly
18 as being a possible source for this temperature
19 excursion .
20 From consulting witha number of experts in the
21 field and some pretty rigorous calculations and
22 experience on the unit itself and other. units , it was
23 determined not to be a significant factor in the
24 excursion itself .
25 There was also' a move made on the hydrogen quench
,,�
Certified Shorthand Reporters
f za
� ndolmi 27 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. i P.O.Box 4107•Concord.CA 9452.1-4107
(510)685-6222•Fax (510)685-3829
1 in the fifth bed during this critical seven and a half
• 2 minute period . That ' s literally how . quickly this
3 happened . And because of the magnitude of the heat
4 released by this demethylization operation, the quench
5 also proved to be an insignificant factor into this
6 excursion.
7 Reactor A itself, as far as its condition, the
8 combination of the nitrogen and all other conditions in
9 the unit , the condition of Reactor A was determined not
10 to have any effect or be a factor in this incident .
11 We did verify that the temperature excursion in
12 the fifth bed through thorough engineering analysis
13 could heat the , pipe up that failed up to its failure
• 14oint . So it did confirm that this incident did happen
PPen
15 in a very short period of time .
16 But what was also clear is , from the interviews ,
17 that up until the explosion the operators . in their own
18 minds didn' t believe they had a temperature excursion
19 and didn' t clearly understand the emergency situation
20 they were in .
21 That ' s a ' little bit about the. investigation . Now
22 I ' d like to talk about the corrective actions , the
23 causes .
24 Some of the relevant factors which enter into our
25 decision on recommendations. This failure is
Certified Shorthand Reporters
Z&Rdondhi 28 2321 Stanwell Drixe•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.sox 4107•Concord.CA 94524-4107
i _� (510)685-6222•Fax(510)685-3829
1' unprecedent in the 34-year history of this unit at Avon .
• 2 And this failure also from our investigation is either
3 extremely rare or unprecedent in this kind of unit . And
4 there ' s over a hundred of these in the world.
5 The pipe was in good shape and clearly failed by
6 being subjected to too high temperatures at normal
7 operating pressure .
8 Causes for the high temperature and the failure
9 was that an excursion took .place in the fifth bed of
10 Reactor Number 5 and . that the unit was not .depressured,
11 which is an emergency operating procedure to safely
12 handle this kind of operating upset .
13 Investigation into the excursion, the temperature
• 14 excursion into Bed 5 , is that after the explosion and
15 the damage as the result of the explosion and the rapid
16 uncontrolled depressuring of the unit , that it wasn' t
17 possible to identify conclusively a single event or
18 combination of events specifically.
19 What we did is that we evaluated all of the known
20 possible causes for temperature excursions in the
21 reactor bed . And these were basically winnowed down to
22 a few. of the most probable ones , or possible ones I
23 should say, which included distribution issues , some
24 phase change issues .
25 As I mentioned earlier, you have a vapor and a
Certified Shorthand Reporters
Lxi��O�Pi�IB
29 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. 1 P.O.Box 410;*Concord.CA 94524-410;
-- -------------
- —� (slo)68.5-6222• Fax (510)685-3829
1 liquid entering the top of this reactor . By the time it
• 2 gets partway through Bed 4 , it becomes 100 percent vapor
3 or gas .
4 And also the .fact that the hydrocracking process
5. is subject to, very rarely though, unknown sources that
6 lead to localized hot spots . And that ' s why in the
7 hydrocracking process you have specialequipment to
8 depressure the unit and you have emergency procedures to
9 do so, to control this rare but significant event .
10 As far as not depressuring the unit by the
11 operators , some of the relevant factors were , is that
12 this excursion, this whole event took place over a seven
13 and a half minute period. It happened in the bottom of
14 the reactor in isolation .
15 Typically in the experience of our operators and
16 other operators at other facilities , these temperature
17 excursions usually involve a number of beds . And also,
18 often it involves a number of cycles or waves through
19 the reactor .
20 In other words , as one bed heats up, that
21 temperature gradually flows to the following . bed, and
22 you' ll see a spike . of - temperatures proceed down a
23 reactor . As it comes out of the, bottom of the reactor,
24 it goes into the heat ,. exc.hangers . And that heat is
25 exchanged with the feed which now reintroduces this high
Certified Shorthand Reporters
�iBItdUItPiTlB 3 0 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Boz 4107•Concord.CA 94524-410
-_-- — - - (510)685-6222 Fax(510)685-3829
1 temperature wave to the top of the reactor .
• 2 And in some of the past histories of .this unit ,
3 units will go through a couple cycles before there ' s any
4 damage if they' re not depressured . ,
5 In our experience , there was a single cycle ,
6 single event , and it happened very rapidly, which is
7 very rare i,n the industry. So the key here is , is that
8 this event was outside the experience of our operators .
. 9 Also as I ' ve described, there were .a number of
10 other events that occurred in the same time period which
11 extract- - - or distracted the operators ' attention . And
12 clearly, if you sum it all up, is that until the
13 explosion occurred, they did not realize the emergency
14 situation that had developed .
15 Other pertinent factors are , is that in. the first
16 quarter of 1996 , there was an excursion incident whereby
17 the emergency depressuring procedures were not followed.
18 And . briefly what these procedures are is if any
19 temperature in the reactor exceeds 800 degrees
20 Fahrenheit , you push - - you manually push a control val-
21 - - controller which dumps the pressure on the unit at
22 300 pounds per minute . This is a depressuring design
23 control device which is part of the license on this unit
24 and it ' s installed on all unicrackers in the world .
25 And of course , the reason why depressuring at a
�r ,� Certified Shorthand Reporters
Zandone la 3 1 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC.
P.O.Bos 4107•Concord.CA 94524-4107
(510)685-6222• Fax(510)685-3829
. 1 controlled rate is if you get pressure too fast , as in
• 2 the case that happened during our explosion, you
3 actually damage the reactor, and that in our case a
. 4 large number of the catalyst support grids and
5 distributors in the lower part of all the reactors were
6 damaged by depressuring in an uncontrolled manner .
7 As a result of this incident , it came to the
8 attention of the supervisor of the unit who became
9 alarmed that these emergency procedures had not been
10 followed . He put in place retraining of all the
11 operators to reemphasize the importance of the emergency
12 procedures and the fact that when temperatures exceed
13 800 degrees the unit should immediately be depressured.
• 14 All the operators were retrained and recertified on this
15 procedure in April of 1996 .
16 In addition, the very crew that was on the night
17 of the 21st had a safety meeting in November of 1996 on
18 the subject of emergency procedures .
19 That ' s an overview of the causes that came out of
20 the investigation . Now I ' d like to talk about the
21 improvements that are going to be. put in place .
22 CHAIRMAN DeSAULNIER : John - -
23 MR . MILLER : Yes .
24 CHAIRMAN DeSAULNIER : Excuse me for just a
25 moment . About how much longer do you have , about five,
�4_zandoRe a Certified Shorthand 520-40Reporters
LL))�Q 3 2 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE,INC P.O.Box 4107•Concord.CA 94524-4107
--� (510)68576222•Fax(510)685-3829
1 ten minutes?
• 2 CHAIRMAN DeSAULNIER : I think the Board
3 appreciates the depth of going through - -
4 MR. MILLER : Four minutes .
5 CHAIRMAN DeSAULNIER : Okay. May we make a
6 suggestion to Mr . Barry, and if it ' s okay with the
7 district representative .
8 There are some people here in regards to Lucy
9 Lane . Maybe you could make yourself available in 108 so
10 you could share the information that we ' ve gotten from
11 the developer, and maybe Marshall could make himself
12 available , so that when it comes back to the Board those
13 folks can decide whether they want to stay or not .
14 And it looks like we ' ll be 20 - - 20 t.o 30 minutes
15 more on this item.
16. Go ahead, John .
17 MR . MILLER : Okay.
18 CHAIRMAN DeSAULNIER : So just one last time - -
19 MR . Did you say Room 108?
20 CHAIRMAN DeSAULNIER: Room 108 , right across the
21 hallway here . If you go right outside here - - anyone
22 who' s here on the Lucy Lane item, which I believe is
23 Item D . 10 , which is at 3 : 00 o' clock, staff and the
'24 developer will be in the room to explain what ' s happened
25 in the last two weeks . We ' ll give those folks a second .
Certified Shorthand Reporters
�ZandoRdIB � 33 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Bos 410-•Concord.CA 94524-4107
(510)685-6222•Fax (510)685-3829
1 Sorry to interrupt .
• 2 MR . MILLER: That ' s okay.
3 CHAIRMAN DeSAULNIER: Go ahead . l
4 MR . MILLER : Okay. Those are the relevant causes
5 and some of the conclusions . And these were used to
6 develop the improvements to prevent this from happening
7 again .
8 First I ' d like to point out that because of the
9 magnitude of this incident , we ' ve approached the
10 solution to it in a very conservative way. In other
11 words , many of the actions that were taken normally
12 wouldn' t be adopted or not. common practice throughout
13 the industry, but because we feel that not 'only do we
14 need to correct thisP roblem to make sure that it never
15 happens again, we also need -to .regain the confidence of
16 the operators and many of the people - involved with this
17 unit , to make sure that they understand that we ' re doing
18 the . utmost to ensure the safety of this unit in the
19 future .
20 The improvements are really divided into two
21 categories , one is mechanical and the other is people .
22 Under mechanical , what we are planning to do and what we
23 are doing is that we ' re replacing all the
24 instrumentation with a state-of-the-art DCS system,
25 distributed control system, which is the standard in our
ZandoneIIa I-1 Certified CA 9 Reporters
3 4 2321 Stanwell Drive•Concord.d.CA 94520-4805
REPORTING SERVICE.INC. P.O.Box 410;•Concord.CA 94524-410
----- -- ----- -- (510) 685-6222• Fax(510)685-3829
1 refinery.
• 2 We ' ll be replacing all the board . mounted
3 temperature data loggers and IA systems with the latest
4 Foxboro generation control system. In addition, we will
5 be installing state-of-the-art distribution trades in
6 the second stage of the hydrocracker .
7 And both .of these are to address all the possible
8 causes that were identified to the possibility of the
9 excursion . But as 1 pointed out , after the explosion we
10 couldn' t narrow it down to any one , so we have addressed
11 them all .
12 In addition to this , because of the very quick
13 time frame and the likelihood that a complex unit like
• 14 this can dilute an operator' s attention, we are going to
15 install an automatic depressuring control .
16 To put things in perspective , of the hundred-plus
17 units in the world, there ' s only a small handful that
18 automate this depressuring control . In our case , we
19 feel it ' s justified.
20 And also all of the temperature indicators that
21 are in the reactor beds , this is. 96 per reactor, are
22 going to be brought into the control room, into the new
23 DCS system .
24 Now, on the people side , we are going to be
25 retraining all of our operators with th=is units on the
�MndoneHa Certified Shorthand Reporters
■ 3 5 2321 Stam+ell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4107•Concord.CA 94524-4107
(510)685-62.22•Fax(510)685-3829
1 emergency procedures , making sure that it is .understood
2 that emergency procedures are meant to be black and
3 white , they are not discretionary. We ' ll also carry
4 this through to our other units .
5 And also we feel that it ' s important to retrain
6 the operators more on some of the fundamentals of how
7 the reactors behave under different kinds of upsets .
8 Aside from bringing these improvements and others
9 as outlined into our report at Avon, we will also be
10 applying those to other refineries in TRC . We have four
11 hydrocrackers within TRC . They' re all unicrackers to my
12 knowledge .
13 I ' m not going to say that we' ll adopt all of
14 these recommendations at all of o.ur' hY drocrackers ,
15 because as I pointed out really earlier, each
16 hydrocracker is a different design, slightly or greatly,
17 and they all behave differently. So we will be adopting
18 the parts that are justified in those other units to do
19 what we need to do is ensure the safe operation of those
20 units .
21 Further, we -will be sharing information from our
22 investigation with other refiners , both locally here in
23 the Bay Area and in the industry as a whole . And at
24 some point we ' ll be working through the licenser to make
25 sure this information gets to all the unicracker
-Certified Shorthand Reporters
�Zandortd 36 2321 Stamwell Drive•Concord.CA 94520-4805
REPORTTNG SERVICE.INC. I Y.O.Box 410;•Concord.CA 94524-4107
(510)685-6222 Fax(510)685-3829
1 licensers in the world .
• 2 That' s basically my review on the unit
3 description, the investigation and the corrective
4 actions .
5 We recognize we need and have begun a major
6 rebuilding process at Tosco, not just the hydrocracker
7 equipment as we ' ve discussed here , but the confidence of
8 our workers and of the public . This includes rebuilding
9 relationships and trust that some people say may have
10 suffered over the past few years .
11 I commit to you that we will operate our
12 refineries safely and responsibly. And we will also
13 work cooperatively with the County.
• 14 Thankou ver much.
Y Y
15 CHAIRMAN DeSAULNIER : Any questions of
16 Mr . Miller?
17 John, before you leave. - -
18 MR . MILLER : Yes .
19 CHAIRMAN DeSAULNIER : - - I have a couple
20 questions , and then Supervisor Rogers .
21 On page 36 of your report , your 12 bullets under
22 recommendations and you just talked about some of them,
23 do you have a time line when you might be finished with
24 those on your recommendations , your corrective actions?
25 MR . MILLER : We will be completing all the
Certified Shorthand Reporters
�Zaildoneffil 37 2321 Stanwell Drive•Concord,CA 94520-4808
I REPORTING SERVICE.INC. P.O.so.4107•Concord.CA 94524-4107
(510)685-6722• Fax (510)685-3829
. 1 . recommendations before the unit starts up in July.
2 CHAIRMAN DeSAULNIER : All right .
3 MR. MILLER : Of course , the only exception might
4 be, is the extent of which. we ' ve basicallyshared this
5 information throughout the industry.
6 CHAIRMAN DeSAULNIER: And the difficulty - - and
7 you and I talked a little bit about this yesterday. The
8 excursions and going over the 800-degree threshold, and
9 I ' m looking at page 27 just before the explosion when - -
. 10 MR . MILLER : Yes .
11 CHAIRMAN DeSAULNIER : - - Bed 5 rose to 1200 and
12 then back -to 1220 and defaulted to zero, but you' d had
13 problems with it before .
14 Now, the automated system, every time it goes to
15 80.0 , it will depressure automatically. Is that what
16 you' re telling us?
17 MR . MILLER : The automatic system will be _
18 designed when. any temperature exceeds 800 degrees , . that
19 the unit will depressure automatically without operator
20 intervention .
21 CHAIRMAN DeSAULNIER : And what kind of cost
22 associated for the refinery is there when that
23 depressurization goes through? You mentioned that
24 things happen when you have to depressure .
25 MR . MILLER : Well , when you depressure , you' re
qq,,,,,�
iii Certified Shorthand Reporters
ezandoneUck 38 2321 Stanwell Drive•Concord.CA 94520-4808
P.O.Box 410;•Concord.CA 94524-4107
REPORTING SERVICE.INC.
R-- (510)685-6222•Fax(510)685-3829
1 basically shutting the unit down . And to reheat and
• 2 repressure the unit - - Chuck, you probably know better
3 the time line to restart the unit after depressuring .
4 MR. - WAITMAN: Well , you' d probably be looking at
5 anywhere from a one to five-day outage and hundreds of
�6 thousands of dollars for each of those days , sort of
7 depending on the circumstances .
8 CHAIRMAN DeSAULNIER : So the question would be is
9 there pressure- on the operators or will there be
10 pressure somehow in the future to avoid that when it
11 happens . I know there ' s - - you' re going through
12 management pressure to make sure it doesn' t happen, but
13 when it does happen - -
14 MR . MILLER : Well , let me make sure it ' s clear to
15 everybody, is that as with any business , there' s always
16 pressure to operate and be in the market . In other
17 words , you don' t make money unless you' re operating .
18 CHAIRMAN DeSAULNIER : Understood .
19 MR . MILLER : But at our facility - - and you
20 basically have my commitment - - we do not . operate
21 equipment unless it can be operated safely. So there is
22 no pressure .
23 Our operators all are empowered that if an unsafe
24 condition develops in their unit , that they can shut
25 that unit down .
Certified Shorthand Reporters
riga tdoneIkk 39 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. Box 410 •Concord,CA x4524-410
(510)685-6222• Fax(510)685-3829
1 CHAIRMAN DeSAULNIER : And in your report , and I
. 2 don' t have the page number, there was a comment about
3 some of your operators being told of the possibility of
4 being terminated if they didn' t follow the procedure in
5 terms of shutting down after 800 . Was that right , did I
6 read that somewhere?
7 MR . MILLER : Well , that - - Chuck, once again, you
8 were in the interview. Was that a quote?
9 MR. WAITMAN: That was a quote from one of the
10 operators , yes .
11 CHAIRMAN DeSAULNIER : And there was never any
12 pressure the other way that you' re aware of on the
13 operators to perhaps not depressure when it goes to 800 .
14 I didn' t see an thin in that regard in the report?
Y g. g P
15 MR . MILLER : To my knowledge , and I asked that
16. very same question too, Mark, absolutely not .
17 CHAIRMAN DeSAULNIER : Okay .
18 MR . MILLER : No operator was ever pressured that
19 in an unsafe condition that they were not allowed to
20 depressure that unit per the procedures .
21 CHAIRMAN DeSAULNIER : All right . Thanks .
22 Supervisor Rogers had a question . .
23 SUPERVISOR ROGERS : The temperature control which
24 went too high and defaulted back to zero, I was - - I was
25 curious about why that couldn' t be built in a way where
Certified Shorthand Reporters
ZaladoItPiljB 40 2321 Staf1N'el] Drive•Conc•ord,CA 94520-4808
REPOMNG SERVICE.INC. P.O.Box 410;•Concord.CA 94524-4107
(510)685-6222�__9 Fax(510)685-3829
. 1 the temperature would be shown as being over the limit .
• 2 I believe it was an 800 figure .
3 In other words , why - - why wasn' t it possible to
4 build that temperature gauge in a way so that instead of
5 defaulting back to zero when it goes off the charts , it
6 goes to a setting that indicates that it ' s too high, not
7 back to zero?
8 MR. MILLER : Well , actually, Chuck, you can
9 probably talk better to that subject .
10 MR . WAITMAN: There was actually two displays as
11 the temperature rose , basically means they' re back to
12 the limit . One was the strip chart indications . And as
. 13 they reached 800 , that was the top of their scale and
• 14 they stayed at 800 . . And it basically indicated an
15 off-range high.
16 The other was a signal that came into the
17 multi- - - came into the data logger . And they came in
18 through a transmitter that we refer to as a multiplexer . .
19 And through that device there was a single error message
20 generated . So any signal that was out of range or
21 failed thermocouple or any other disturbance with that
22 multiplexer, the only signal that could transmit was a
23 zero . . And so it transmitted that zero signal to the
24 multiplexer when you were out of range .
25 And yes , there are - - you know, there ' s numbers
Zandonell8 Certified Shorthand Reporters
j 41 ?321 Stanwell Drive•Concord,CA 94520-4508
j REPORTING SERVICE.INC. P.O.Box 4107•Concord.CA 94524-4107
�_- — (.510) 655-6222• Fax (.510)685-3829
.1 of different types of hardware on the market , but that
2 was the limitation of that device that was installed.
3 MR . MILLER : Also Supervisor Rogers , bear in mind
4 that with the emergency procedures that were in place ,
5 rarely, if ever, should a temperature have approached
6 1400 degrees and defaulted to zero .
7 CHAIRMAN DeSAULNIER: Thanks , John .
8 Supervisor Gerber has some questions .
9 SUPERVISOR GERBER : Yes . , Thank you very much for
10 your presentation .
11 As you know, the County' s interest has been
12 clearly in terms of this concept of root cause analysis .
13 And I am pleased to see that , I think at some level , .
14 hopefully we ' re on the same page here in terms of the
15 importance of really honing in on where are the systems
16 that somehow lead to, in this case, an accident . And I
17 just wanted to ask you a couple of questions about - -
18 about your presentation in that regard .
19 You mentioned that in April of 196 there was an
20 excursion - -
21 MR MILLER : I believe it was in March . Wasn' t
22 it , Chuck?
23 ' MR . WIGHTMAN : Yes .
24 SUPERVISOR GERBER : Okay. March of -96 , an
25 excursion where the temperature went over 800 degrees
�Zq,,� Certified Shorthand Reporters
aladoRCI 8 42 2321 Stanwell Drive•Concord.CA 94520-4808
P.O.Box 410 •Concord.CA 94524-410 i
REPORTING SERVICE.INC.
_(510)685-6222•Fax(510)685-3829
1 and depressuring was not done at that time,, -correct?
2 MR. MILLER : That ' s correct .
3 .SUPERVISOR GERBER : Okay. And, but apparently
4 whatever was done , it worked.
.5 MR. MILLER : It worked, but it basically wasn' t
6 in accordance with the emergency procedures - -
7 SUPERVISOR GERBER : Right .
8 MR . MILLER : - - which is - - I hope you
9 understand, emergency procedures needto be black and
10 white . Because if you have judgment , then you don' t
11 really have an emergency procedure .
1.2 SUPERVISOR GERBER : Uh-huh . If - - I ' m trying to
13 figure out a way to ask you this without putting you on
14 the spot in particular .
15 Clearly, what you' ve described, and I think also
16 what the Health Departent described from the County, is
17 that whatever occurred that particular evening, part of
18 the difficulty for the operators was that it was out of
19 the ordinary .
20 MR . MILLER : Yes .
21 SUPERVISOR GERBER : In other words , not
22 consistent with their experience , so though some of them
23 had had ten years of experience or whatever, what was
24 going on there was not consistent with their experience ,
25 made it more difficult for them perhaps to predict or
M�r,,� r Certified Shorthand Reporters
nikone lik 43 2321 Stanwell Drive•Concord.CA 94520-4808
REPOKI I2VG SERVICE.INC. P.O.Bos 4107•Concord.CA 94524-4107
(510)685-6222• Fax(510)685-3829
. 1 make judgments about what to do?
• 2 MR . MILLER : Or might I had, anybody' s experience
3 at Tosco .
.4 SUPERVISOR GERBER: Okay. So if we ' re going to
5 prevent this from happening in -the future, it seems to
6 me - - correct me if I ' m wrong - -that there has to be a
7 clear direction within the company that not only do you
8 have procedures that say, or even equipment that perhaps
9 can be automated to include that when the temperature
10 goes over 800 , automatically we depressure , but wouldn' t
11 it - - wouldn' t you also need basically an instruction
12 that says when you see something that .you don' t
13 understand or that is- outside your experience , then err
14 on - the side of safety and depressure . Is that - -
15 MR . MILLER : And that is exactly what we will be
16 emphasizing in this case in particular .
17 SUPERVISOR GERBER : The other question I wanted
18 to ask you aboutis the - - and again, I ' m trying to
19 catch up' with your technology, so if I kind of mush all
20 over it , feel free to correct me .
21 But the reactor in question was built in the
22 ' 60s?
23 MR. MILLER : That ' s correct .
24 SUPERVISOR GERBER : And I understand that the
25 part of what you think influenced the hot spot has
r Certified Shorthand Reporters
�7i8IadUIt6$ 44 2321 Stanwell DriNe•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4107•Concord.CA 94524-4107
� — — — — INC. (510)685-6222• Fax (510)685-3829
1 something to do with the distributors?
• 2 MR. MILLER : That ' s one of the possibilities that
3 was evaluated .
4 SUPERVISOR GERBER : Okay. When was the last .time
5 . that the distributors, or if there' s more than one , in
6 that reactor were redesigned to current technology, do
7 you know?
8 MR . MILLER : I don' t know. Rick, can you comment.
9 on that?
10 MR . BONNER : There have been modifications
11 over - - there have been modifications done over the '
12 years . I know in January of 196 , there was a thorough
13 inspection of all the distributors in that reactor, and
14 they were put back in before the catalyst exploded .
15 I don' t off the top of my head recall the history
16 of the improvements that were made to those
17 distributors .
18 MR . MILLER : This by the way is Rick Bonner . He
19 is the health and safety environmental manager for the
20 Avon refinery and the Rodeo refinery.
21 SUPERVISOR GERBER : Hello .
22 Well , if I understood your plans for correction,
23 you are going to redesign the distributors , correct?
24 MR . MILLER : What we ' re doing is we ' re installing
25 state-of-the-art design distributors , and really for two
�Zq� Certified Shorthand Reporters
ando�PiIIB 45 2321 Stanwell Drive•Concord.CA 94520-4808
I REPORTING SERVICE.INC. I P.O.Box 4107•Concord.CA 94524-4107
(510)685-6222•Fax(510) 685-3829
1 reasons . One is , is that to a large extent the existing
• 2 distributors as a- result of the explosion were damaged
3 to destroyed, so that they need replacement anyway.
4 And since we have to replace a large number of
5 the distributors , we made the decision to go with
6 state-of-the-art design .
7 SUPERVISOR GERBER: Okay-
8 MR. MILLER : And in, addition, as I said earlier,
9 is that the new distributor trays solve many of the
10 possibilities that were identified as contributing to
. 11 the temperature excursion .
12 SUPERVISOR GERBER : So if I - - if I get it ,
13 that - - that mechanical change may in fact deal with the
• 14 underlying root cause of what caused the hot spot?
15 MR . MILLER : Well , depending on how you define
16 the. term. But it was definitely - - it is definitely a
17 solution to many of the factors that were possible
18 contributors to the temperature excursion in both Bed 4 -
19 and Bed 5 of Reactor 3 .
20 SUPERVISOR GERBER : And in addition, the - -
21 you' re moving the controls that Mr . . Glanzman went out
22 to he went outside to .look at temperature controls or
23 whatever . Those are being moved inside the control room
24 so that an operator in a similar circumstance would not
25 have to go outside?
Certified Shorthand Reporters
;1 ZaradoneIIa 4 6 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Bo 4107•Concord.CA 94524-410-
(510)685-6222•Fax(510) 685-3829
1 MR . MILLER : I need to back up . Actually, this
• 2 was .something, as I said, was planned in the phase
3 approach _ to happen anyway, but something I guess I
4 didn' t make clearer when .I was discussing operator
5 duties .
6 Part of the safe operation of a unit involves
7 operators who are assigned both inside and outside . It
8 is the normal duties of all operators to routinely tour
9 their units . In other words , we rely on operators to
10 use their senses , their sight , their- sense of smell ,
11 their hearing, their touch, to monitor our equipment .
12 And the situation relative to the reactors are,
13 is that they normally operate it , are very safe , and
14 there are no risks in being in their proximity during
15 normal operation .
16 So I do want to emphasize that it is common
17 practice throughout industry, and as a matter of fact we
18 . absolutely require our operators to be in the units as
19 part of the safe operation of all facilities , not only
20 in our facilities but any facility in the world.
21 SUPERVISOR GERBER: Okay.
22 CHAIRMAN DeSAULNIER : Supervisor Rogers has a
23 question .
24 SUPERVISOR ROGERS : I had a question about
25 previous comments or suggestions , complaints , ,whatever .
Z,,� Certified Shorthand Reporters
and �QillB '—� _ 47 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC i P.O.Box 4107•Concord.CA 94524-4107
(510)685-6222•Fax(.510) 685-3829
1 Are you or other - - I guess I ' m a little bit unclear
• 2 about what complaints if any there had been or
3 suggestions for improvement . there have been for
4 employees prior to: this that you or other Tosco
5 management was aware of connected with this -
6 MR. MILLER : Can you be more specific?
7 SUPERVISOR ROGERS : Well , just were there
8 comments or suggestions about this hydrocracker unit and
9 suggestions that you should change something or put it
10 on automatic or, you know, do something different with
11 it .
12 Have you had that kind of suggestion about the
13 unit which exploded?
• 14 MR. MILLER : You know, I - - I ' m trying to
15 understand what you' re saying . I ' m sorry.
16 SUPERVISOR ROGERS : Okay. Let me - -
17 MR . MILLER : I ' m still struggling with - -
18 SUPERVISOR ROGERS : Let me - -
19 MR. MILLER : - - specifically what you' re asking .
20 SUPERVISOR ROGERS : Okay. Let me try it again .
21 Are you aware of management having had
22 suggestions for changes in operational procedures from
23 employees prior to this incident , from those employees
24 who were familiar with it and were working with it?
25 MR . MILLER: As far as operating procedures?
� * Certified Shorthand Reporters Zaildom la 48 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. j P.O.Box 4107•Concord,CA 94524-4107
(510)685-6222•Fax(510) 685-3829
1 SUPERVISOR ROGERS : Correct .
• 2MR . MILLER : I ' m personally not aware . I
3 wouldn' t find it surprising in that in addition to, as
4 the unit goes through the years of upgrading the
5 equipment , upgrading your procedures is a normal part of
6 improving the safety of any unit .
7 And as a matter of fact , we solicit suggestions
8 from all involved, including the operators , the
9 mechanics , the engineers, anybody. And these
10 suggestions come in on all the units .
11 Now, on this particular unit , I don' t know - -
12 Rick, could you comment as far as any suggestions
13 relative to the Number 3 reactor?
• 14 MR . BONNER : I ' m not aware of anything specific
Y g P
15 to the Number 3 reactor . It ' s very common for the
16 operators to have suggestions as part of looking at
17 improvements in the units . There are a number of
18 projects that are always on the - books (unintelligible)
19 employees to consider (unintelligible) .
20 SUPERVISOR ROGERS : Thank you . And I do want to
21 thank you also for the cooperation which has been
22 extended . The Board as you know was very concerned that
23 Dr . Walker would be able to look us in the eye and say
24 he ' s had the access he needs to do a proper
25 investigation, and he has assured us of that .
Certified Shorthand Reporters
�ZdndonoIla 49 2321 Stanwell DriN e•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4107•Concord.CA 94524-4107
(510)685-6222•Fax (510)68.5-3829
1 And I think I speak for the Board in noting that
• 2 we appreciate that access and the ability to know that
3 not only we can benefit from your report but from
4 Dr . Walker' s report .
5 Theother question I had which is a concern is
6 that we have several employees here who are , I guess at
7 its most basic level , looking at something where they' re
8 kind of scratching their heads a little bit and they
9 don' t really understand it exactly.
10 And they' re - - the picture I get is that they' re
11 concocting theories about why it is that everything is
12 okay . And it turned out very tragically that the
13 theories were wrong.
• 14 I guess the question I ' m wondering about is how
15 often does this happen in other parts of your- refinery' s
16 operation, not necessarily the hydrocracker, but how
17 often is it that your front line personnel are looking
18 at something and they' re looking at some gauge which
19 shouldn' t say such and such and they' re saying,, "Well ,
20 you know, I know this is kind, of unusual but I ' m going
21 to let it go and, you know, continue to monitor it , I ' m
22 not going to do something dramatic , I ' m not going to
23 shut it down . "
24 How much does this kind of situation come up?
25 MR . MILLER : Well , in my personal experience ,
- �'LK1r�^ Certified Shorthand Reporters
llikolmiit a , 50 2321 Stanwell Drive•Concord.CA 94520-4808
P.O.Box 4107•Concord.CA 94524-410
REPORTING SERVICE.INC.
(510)685-6222•Fax (510)685-3829
1 very rarely. But this is something that Rick Bonner can
• 2 answer because he was operations manager far longer than
3 I ever was .
4 MR . BONNER : The operators' job, as John
5 mentioned earlier, is to check and very the operation of
6 equipment . It ' s very routine for them to see a reading,
7 if they question it , they' ll take a reading outside and
8 compare this information .
9 A lot of our. control systems have various levels
10 of redundancy built in so that they have access to that
11 information .
12 I think the important thing to understand in this
13 event is that the operators were actively trying to
14 understand what was going on in the unit . The operating
15 regime that they were in was totally different than any
16 experience than any of us had had in the Avon refinery,
17 and it happened very, very quickly.
18 So the operators were busy trying to understand
19 it and indeed trying to manage the situation to the best
20 of their ability. But it happened so fast and was so
21 unusual that they did not (unintelligible) .
22 SUPERVISOR ROGERS : Thank you .
23 CHAIRMAN DeSAULNIER : Thank you, Mr . Rogers .
24 Ms . Gerber .
25 SUPERVISOR GERBER : Do you know what the
Certified Shorthand Reporters
kR&EIPORTING
RdORCIII, 51 2321 Stan ell Drise•Concord,CA 94520-4808
SERVICE.Itic. P.O.Roz 410 •concord.CA 94524-410
-- _-� (510)685-6222 0 Fax (510)685-3829
1 frequency is of these - - these temperature excursions
2 which would require emergency depressuring?
•
3 MR . MILLER : In our unit?
4 SUPERVISOR GERBER : In your unit , in your
5 company, in other companies .
6 You said that you had - - one of the things you
7 had done was research the hydrocracker reactors to see
8 what sort of is the experience with them as a -
9 MR . MILLER : Okay. In our unit it ' s very rare .
10 Chuck, you can probably talk to what you found
11 when you did your survey, or your group did the survey.
12 MR . WAITMAN : I think there ' s a - - you know,
13 there ' s a variety of experiences . I think there ' s units
• 14 in the country that have never seen an excursion, and
15 there ' s probably temperatures in the unit that have seen
16 excursions on the frequency of maybe one or more a year .
17 And County staff I think has confirmed some of these
18 things as well .
19 SUPERVISOR GERBER : I guess my concern is , and
20 I ' m just - - I ' m sort of harping on this , but it kind of
21 gets back to what Supervisor Rogers was saying .
22 It ' s inherently the case that , at least in my
23 experience , that employees who are trying to do a good
24 job and do their part to assist in the mission of the
25 company, et cetera, will - - will attempt to compensate
r�r,,� ri Certified Shorthand Reporters
J ZaiadU�7�CEI-N�C
tPill ', 52 2321 Stanwell Drix e•Concord,CA 94520-4808
REPORTING S . P.O.Box 4107•Concord,CA 94524-4107
(510)685-6222• Fax(310)685-3829
1 for or do whatever they need to do in order to keep
2 things going .
3 And that unless there is really a very proactive
4 stance from the top down that says - - that essentially
5 gives them permission to depressure , go off line , you
6 know, whatever the phrase is, that their tendency is to
7 keep things going and keep things working .
8 And I realize that what you' re- saying now is that
9 you have - - that you will make it clear that the policy
10 to depressure if =- if going over 800 degrees , that you
it will make it clear that that policy is an absolute , that
12 it gets adhered to. no matter what , right?
13 MR . MILLER : That ' s correct .
• 14 SUPERVISOR GERBER : Okay. And I asked you
15 earlier if you would also make it a policy that when in
16 doubt , depressure .
17 MR . MILLER: That is in place and that will be
18 reinforced .
19 If I could take two minutes , let me give you an
20 example , not this unit but what occurred at the refinery
21 almost two months ago .
22 After the turnaround at one of our crude units ,
23 one of the operators spotted a small flame off the - -
24 off the side of a fractionator . A fractionator looks
25 somewhat similar to a reactor, just the insides are
Certified Shorthand Reporters
Zandond a 53 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4107•Concord.CA 94524-4107
(510)685-6222•Fax(5 10)685-3829
1 different .
• 2 That operator thought that he had a small leak or
3 crack in that vessel and shut the unit down. He didn' t
4 have to ask permission because that is his
5 responsibility and they have the authority.
6 We stripped the insulation in the proximity of
. 7 where the fire was spotted. And we looked and we looked
.8 and we looked . We pressured up the tower with steam.
9 We inspected all around the tower . We couldn' t find any
10 thin spots .
11 So after about two days, and this was two days of
12 lost production, we. made the decision that maybe it was
13 some phenomenon where it took the right temperature
14 o eratin temperature and pressure, that there was some
P g
15 kind of . microscopic crack and that ' s what would happen,
16 and it would open up and we ' d spot it and we could make
17 repairs , because we didn' t know where it was .
18 We started up the unit . There was no leak . Our
19 best theory is , is that during the turnaround, some
20 small amount of oil had leaked behind the insulation on
21 the vessel and had gradually migrated down - - you know,
22 normally on a fractionator, the top of the tower .is hot
23 and the bottom of the tower is very - - I mean the top is
24 cold and the bottom is very hot . And it finally got
25 down to a zone to where it was hot enough to flash .
i Certified Shorthand Reporters
�itdUItPiilB
54 2321 Stanwell Driwe•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 410;•Concord.CA 94524-410;
(510)685-6222•Fax(510) 685-3829
1 There was no leak on that tower .
• 2 But I will put to you that that operator. made the
3 correct decision and that - - if that happened today, we
4 would shut that unit down again .
5 CHAIRMAN DeSAULNIER: John, you' ve been up
6 there - - up here for a while - and I see Jim wants to
7 ask another question - - , looks like 35 , 40 minutes.
8 But I piggyback on what Supervisor Gerber asked
9 in terms of the number of excursions and how many times
10 you' ve depressured .
11 On pages 18 and 19 of your report , it says that
12 you had four excursions that exceeded 800 degrees
13 Fahrenheit after November 196 when you had a safety
14 meeting going over the guidelines and the procedures .
15 And then on 19 - - so that ' s four excursions just since
16 November ` 96 .
17 And then the quote, I ' ll read, quote : "Many of
18 the operators reported that they have experienced
19 numerous temperature excursions , but most could recall
20 only one instance when the unit was depressured as a
21 result . "
22 And then it goes on, the quote I was looking for
23 before : "One operator reported that he had been told
24 that he could be fired for failing to be - - to
25 depressure when necessary in the future . "
� Certified Shorthand Reporters T,aradoAmIla —� 55 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. j P.O.Bos 4107•Concord.CA 94524-4107
(510)685-6222•Fax (510) 685-3829
1 So I guess I ' m a little concerned that -- and
2 Chuck, you said maybe one excursion a year . But in this
3 report , your report , it says that there were four since
4 November .
5 Am- I not reading this right or am I - taking it out
6 of context?
7 MR. MILLER : Chuck - -
8 CHAIRMAN DeSAULNIER : And what I - -
9 MR. MILLER : Chuck, you want to talk .to that?
10 That came out of the interviews in the investigation,
11 did it not?
12 CHAIRMAN DeSAULNIER: Just before you - - I guess
13 m,y concern is I think what Supervisor Gerber was after,
• 14 how many excursions over 800 degrees took place - -
15 correct me if I ' m wrong - - and as a . result , " I guess
16 we ' ve only depressured once .
17 So, Chuck, help me .
18 MR . WIGHTMAN : I think you know of - - the report .
19 refers to the March 19th and January 19th event where we
20 believe - - you know, where we believe the temperatures
21 went over 800 .
22 The other events, we do believe were due to, I
23 think there' s three other that are forecast there, were
24 due to temperature indications that were one way in
25 another error, which is one of the things that - -
�r,,� Certified Shorthand Reporters
56 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTINIG SERVICE.INC. P.O.Box 4 107•Concord,CA 9452 -4107
(510)685-6222• Fax (510)685-3829
1. CHAIRMAN DeSAULNIER : You have to correct .
2 MR . WIGHTMAN: - - Supervisor Gerber is getting - -
3 CHAIRMAN DeSAULNIER: Right .
4 MR. WIGHTMAN: - - getting at .
5 MR. MILLER: But I also want to emphasize that ,
6 unfortunately the excursion that occurred on the 19th
7 went undocumented . And therefore, supervision was not
8 aware that the procedure had not been followed until the
9 investigation as a result of the 21st .
10 CHAIRMAN DeSAULNIER : All right .
11 MR . MILLER: And this is an issue that is also
12 being addressed, moving forward to ensure that
13 documentation is there , so that we can better ensure
• 14 that procedures are followed.
15 CHAIRMAN DeSAULNIER : Jim, you had a question?
16 SUPERVISOR ROGERS : You had mentioned that
17 there ' s - - there will be a change with the unit being
18 automatically depressurized, shut down, if the
19 temperature hits a certain degree .
20 And my understanding - - I just want to make sure
21 I understand this right is that if that had been in
22 place prior to this accident , we would not have had the
23 accident . Am I understanding that correctly? Or at
24 least it would have been extremely unlikely we would
. 25 have had the accident .
Certified Shorthand Reporters
ZalRdorad —1 57 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTIrIG SERVICE.INC. P.O.Box 4107•Concord.CA 94524-4107
(510) 685-6222• Fax(510)685-3829
1 MR . MILLER : That ' s correct . If the unit. had
2 been depressured, it ' s not likely that the pipe would
3 . have ruptured .
4 SUPERVISOR ROGERS : And what was - - was the idea
5 of having an automatic shutoff machinery rather than
6 relying on people to do it , had that idea been discussed
7 or talked about or considered?
8 I think I heard you earlier saying that almost
9 all similar units throughout the country do have that
10 automatic feature , had that - -
11 MR . MILLER : No, you misunderstood me . The vast
12 majority of all the hydrocrackers in existence do not
13 have automatic depressuring devices .
• 14 A companies large number of all over the world
15 have looked at this issue . And the vast majority of
16 them have come to the conclusion that , through the
17 procedures , that this is adequate safety measures .
18 Only a very few companies have adopted installing
19 automatic depressuring universally in their company.
20 And so this is a company-by-company basis .
21 But one thing I need to emphasize is that what we
22 present to you today is one type of hydrocracker . There
23 are many, many different designs . I mean they all do
24 the same thing but , you know, as there ' s many, many
25 different cars , you know, and they all behave
Z��►r^� Certified Shorthand Reporters
andondhk 5$ 2321 Shjmrell Drip e•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Bo.4107 0oncord.CA 94524-4107
(510)685-6222•Fax(510) 685-3829
1 differently.
• 2 And so what 'is necessary to operate one facility
3 safely isn' t necessary for another . But we at SFAR at
4 Avon have made the decision . that we feel that this is
5 necessary for this unit .
. 6 SUPERVISOR ROGERS : What - - what types of
7 activities are being done at the Rodeo refinery? I
8 understand that one part of this process is really
9 trying to reemphasize certain points about how people do
10 their job and the tradeoff between, you know, taking the
11 extra step for safety. And I think all that seems
12 helpful and important .
13 To what extent will that be going on at the Rodeo
• 14 refinery as well as the Avon refinery?
15 MR. MILLER : We are taking all the lessons
16 learned at Avon and applying them to Rodeo . An engineer
17 has been assigned to review the operation of the Rodeo
18 hydrocracker .
19 In addition, within TRC, we' re sharing that same
20 information . Very shortly we will be sharing that
21 information with all the refiners in the Bay Area, and
22 ultimately all the licensers in the world, through the
23 licenser, of course .
24 SUPERVISOR ROGERS : Thanks .
• 25 CHAIRMAN DeSAULNIER : All right . Is that the end
qq^�
Certified Shorthand Reporters
Zandonena 59 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC.
P.O.Box 4107•Concord.CA 94524-4107
(.510)685-6222 e Fax (510)685-3829
I of your presentation, John?
• 2 MR . MILLER : Unless you' ve got another question,
3 Mark . .
4 CHAIRMAN DeSAULNIER: No, I don't . But I did
5 want to say, in my experience here on the Board, we
6 frequently have had requests to the refineries to come
7 and .explain situations to us . And this is the most
8 information that' s ever been afforded to me , and I
9 appreciate that.
10 It raises some concerns , but I appreciate being
li able to have concerns and questions at least asked .
12 Because in the past we ' ve gotten a lot of from. other
13 refineries , "No, that ' s privileged information, we' re
• 14 not going to give it to you . "
15 So we do appreciate that change .
16 MR. MILLER : Okay. Thank you .
17 CHAIRMAN DeSAULNIER : Dr . Walker?
18 DR . WALKER : Yes , Mr . Chairman . I would be the
19 last to say that this is not rocket science . But I
20 would like to just take a few minutes to emphasize that
21 what root cause analysis is all about is not simply to
22 point at employee error, but to look at the systems , the
23 management practices and the preventive measures that
24 could have helped prevent the accident from happening .
25 I ' d like my staff of chemical engineers , Bill
Certified Shorthand Reporters
ZandUneffi8 60 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Boy 4107•Concord:CA 94524-4107
------ (510)685-6222• Fax (510)685-3829
1 Alton and Laura Brown, to comment briefly on their
• 2 analysis of the root cause of this incident .
3 MR . ALTON: And we will try to make this brief .
4 We started our investigation the day after the
5 incident . And we were along with Cal Osha, the Air
6 Board and Federal EPA, and Tosco provided us offices .
7 Our investigation consisted primarily of
8 reviewing documents provided by - - by Tosco and also by
9 using interviews of Tosco employees . Tosco provided us
10 answers to many written questions . They supplied us
11 operating data, the same data they were working with,
12 temperature data, pressure, temperature and flow charts .
13 We had alarm logs . We had the operator shift
• 14 reports and refinery shift logs . We had the drawings
15 that they were working with, the piping and instrument
16 diagrams that show the details of the reactor .
17 And we had the reactor - - we had the reactor
18 details also . We also had their catalyst data and
19 operating procedures .
20 The interviews that we used, we had written
21 witness statements from both Tosco employees and
22 contractors that were there during the incident . Cal
23 OSHA did most ' of the initial interviewing and gave us
24 briefings on those . -
25 And then subsequently we did our own interviews
�
Certified Shorthand Reporters
�ZBItdUltPiliB 1 61 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4107•Concord,CA 94524-4107
(510)685-6222•Fax (510)685-3829
1 of some Tosco employees and also had discussions with
2 the Tosco investigation team. In the process we also
3 was able to do a close visual examination of the
4 ruptured pipe .
5 our initial conclusion was - - and it wasn' t
6 realize , as Bill says, rocket science - - that the pipe
7 ruptured due to overheating . And we based that on two
8 things , that the temperatures - - the temperature data
9 indicated that the pipe was subjected to very high
10 temperatures .
11 There was a witness a little ways away from the
12 incident that saw this red - - this pipe glowing red at
13 the. time , and so from that we could conclude that . We
14 didn' t- know the reason for the high temperature .
15 From the visual inspection in the field, you
16 could see that the pipe failed in a straight section .
17 It did not fail at any weld or bends . And there was no
18 evidence , visually anyway, of any corrosion, erosion or
19 cracking of the pipe .
20 In fact , you know, the pipe looked to be in
21 fairly basically good shape at that time, but that
22 wasn' t certainly a metallurgical lab exam we did .
23 Subsequent to that it was determined that the
24 pipe was a proper material . It was of a good thickness
25 and good metallurgical properties .
M'^ i Certified Shorthand Reporters
ildUltPH 62 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4107•Concord.CA 94524-4107
(510) 685-6222•Fax (510)685-3829
1 We were after root cause , and we went about that
2 in two ways . We looked at possible ways that this
3 temperature excursion could have been triggered, and
4 then sort .of by process of elimination basically wound
5 up where Tosco ,did.
6 One thing that can trigger this kind of a
7 reaction is if the operators give it a control bounce,
8 maybe a heavy-handed move on the controls . There was no
9 evidence of this at all . The operators were operating
10 ' very skillfully up to the accident .
11 There ' s a machine called a recycle compressor in
12 the plant . . If you lose that , you can have this reaction
13 occur . That did not occur .
14 At one point we were looking at maybe the
15, catalyst bed support had collapsed, putting catalyst
16 where there should not be catalyst . Investigation of
17 the reactor after the incident showed that- did not
18 occur .
19 We looked again at this hydrogen thing that John
20 Miller talked about , sudden reactivation of the
21 catalyst . We beat that 'one to death ourselves , and
22 there' s no evidence that that happened .
23 So we got down to that there was a hot spot down
24 there on Bed 4 that triggered this and probably from
• 25 poor distribution, but we really don' t know.
i Certified Shorthand Reporters
zawkoIt na � 63 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC.
P.O.Box-3107•Concord.CA 94524-4107
(510)685-6222•Fax(510).685-3829
1 We went through a sequence of events , and I ' m_ not
2 going to go through that again. We totally agree with
3 Tosco' s analysis of the sequence of events . That ' s
4 facts you can put together from charts and tables.
5 So what are the contributing causes? One
6 contributing cause, key contributing cause is that the
7 written procedure for shutting down the plant was not
8 followed. But why was that? The operators didn' t
9 believe there was a temperature runaway, and John Miller
10 went into that .
11 The operators had prior instances of problems
12 with this temperature logger and they were - - they
13 didn' t - - they didn' t trust these bouncing readings and
14 the inconsistent readings they were getting at that
15 time .
16 The operators didn' t know that a default value of
17 zero occurs when .you go too high on the temperature
18 scale .
19 And I think evidence of this distrust was that
20 just seconds before the incident , the operators had
21 called for an instrument technician to come in and check
22 out the instrument .
23 A little more subtle thing was this business of
24 the methane . When you have a lot of reaction, you would
25 expect a lot of makeup hydrogen to be coming into the
r ,, Certified Shorthand Reporters
�ZBItdUILPilT8 1 6 4 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING $ERYICE.INC. P.O.Box 410 oncor
•Cd,CA 94524-3829
(510)685-6222•Fax(510)685 38299
1 plant - to the plant to be replacing the hydrogen
• 2 that ' s consumed in the reaction . This wasn' t occurring .
3 In fact , the hydrogen makeup was falling off and
4 eventually went to zero. This was.. because the methane
5 backed it off .
6 They didn' t have - - they had an analyzer, but it
7 was seven minutes time .delay on getting that sample , and
8 so they got an alarm, a low purity alarm, but it was
9 after the incident . Immediately after the incident ,
10 that alarm went off .
11 And again, this - - the temperature increase was
12 incredibly fast . And we couldn' t come up with anything ,
13 in industry that would match that . And certainly it ' s
14 beyond Tosco' s experience .
15 And another thing, they were paying attention _to
16 another furnace at the time .
17 So we get down to root causes here . Our root
18 cause is basically that the pipe - - there ' s two .- - the
19 pipe failed due to excessive temperatures caused by
20 runaway triggered in Reactor - - in Bed 4 and carrying on
21 through Bed 5 .
22 . And then the other basic cause is the operators
23 didn' t trip the emergency system because Tosco may not
24 have adequately trained them - - train them that they' re
25 expected to use it whenever they' re faced with
Certified Shorthand Reporters
�MndoRdhk 65 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4107•Concord.CA 91521-4107
(510)68.5-6222•Fax (510)685-3829
1 conflicting data, questionable data, possibly inaccurate
• 2 data, but that data would normally dictate a shutdown .
3 And that ' s our root cause .
4 CHAIRMAN DeSAULNIER: Laura, you have something
5 to add?
6 MS . BROWN: Just a couple more minutes . Sorry.
7 CHAIRMAN DeSAULNIER; Okay. That ' s fine .
8 MS . BROWN : I ' m going to go briefly through . our
9 recommendations . I ' m not going to go through each
10 recommendation as you have it in your packets .-
11 Our recommendations - - the first thing I should
12 say is our recommendations have all .been accepted by
13 Tosco and they will all be completed.
14 The main recommendation, though, is the automatic
15 high temperature shutdown which John Miller talked about
16 in detail . Tosco is putting that in place, and of
17 course that would shut down the unit if the reactor, or
18 any of the reactors in that particular stage, any of the
19 three , hit the 800-pound maximum operating temperature .
20 And the second main portion of our recommendation
21 was the training of the operators . And I think
22 Supervisor Gerber kind of got right to the edge of it ,
23 you know, how come the operators didn' t shut down when
24 they saw this conflicting information .
25 And that ' s one of our recommendations to Tosco,
Certified Shorthand Reporters
L)Q����PiliB I
66 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC 1 P.O.Box 4107•Concord.CA 94524-4107
(510)685-6222•Fax(510)685-3829
1 of course, is to train your operators, if you have
• 2 something that doesn' t make sense, you' re not sure, you
3 know, go ahead and shut it down, that ' s the safest way
4 to go .
5 I also would kind of like to make the comment
6 here that that is a very difficult culture to get to in
7 industry . And that ' s a big task to ask them to do, and-
8 they said they' re going to do it . So that' s something
9 just to keep in mind there.
10 I wanted to talk briefly about the root cause
11 analysis . And I know Dr . Walker will probably touch on
12 this again .
13 We of course at the Health Department believe
14 that root cause should be connected for all major
15 incidents . Without doing a root cause analysis , we ' re
16 not going to know what truly caused the incident , 'what
17 systems failed, and what we can _do to prevent those
18 incidents .
19 It ' s important to realize that here in 1997 , and
20 in the years to come , the incidents that we ' re having
21 now are combinations of failures . It ' s not typically
22 one failure .
23 And we' re going to have to do some pretty good
24 investigation to find out . what those .combinations are
25 and try to prevent all the way along to make sure that
�r� ■ 1 Certified Shorthand Reporters
�LK1����.PilTB
67 2321 Stanwell Drive•Concord,CA 94520-4808
REPORMG SERVICE,INC. P.O.Box 4107•Concord.CA 94524-4107
(510)685-6222 0 Fax(510)68.5-3829
1 those combinations don' t line up and allow that incident
2 to occur..
3 We have - - our department has already forwarded
4 the information to the local Bay Area refineries ,
5 including Chevron, Shell - - I haven' t seen them here
6 today, but I do have a packet for them, as well as
7 Exxon, so we did include Exxon in that , so they do have
8 this information .
9 I also wanted to make one last comment , and. that
10 is both Bill and I worked a number of months on this
11 investigation . And we feel that Tosco has been pretty
12 candid in their report.,
13 They of course discussed the past temperature
• 14 excursions , which they didn' t have to put in . a public
15 document but they chose to . I think it was the right
16 thing to do .
17 And those excursions certainly did in one way or
18 another contribute in the end to this incident . It
19 allowed the operators to possibly feel that maybe they
20 could, you know, get it to return and they' ve seen it
21 and they' ve been able to pull it back before .
22 _ Unfortunately, in this particular incident there
23 was a lot of confusion about what was happening, and so
24 the operators didn' t recognize this particular incident
25 as an emergency situation .
Certified Shorthand Reporters
ZartdorteIIa '�
68 2321 Stirnwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4107•Concord.CA 94;524-4107
i
(510) 685-6222•Fax(510)685-3829
1 With that , I ' d like to thank the Board for their
• 2 time and patience . I .know it ' s been a rather long
3 presentation ..
4 Dr . Walker . .
5 DR. WALKER: Just one last statement , only to say
6 that we feel that the department has had the opportunity
7 to use not only its authority as a certified unified
8 program agency for hazardous materials in the County,
9 but also representing the public health of the
10 community.
11 And that we feel that the type of presentation
12 that ' s been done today representing a thorough
13 investigation of root cause is really what needs to
14 happen for any future events that happen - to occur . And
15 we appreciate the Board' s support in helping us to
16 achieve that information, that level of investigation .
17 We also acknowledge the position you took this
18 morning in supporting the National Chemical Safety Board
19 in terms of a national approach to this .
20 Thank you .
21 CHAIRMAN DeSAULNIER: All right . Any questions
22 of staff? We do have requests to speak from the public .
23 Supervisor Rogers .
24 SUPERVISOR ROGERS : Could staff go over the
25 differences , if any? It seems to me that in general
Certified Shorthand Reporters
�:liandoneIIa � 69 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Bos 410;•Concord.CA 94524-4107
�— --_Ji (510)685-6222•Fax(510)685-3829
1 it ' s clear that your analysis of this is similar to
2 Tosco' s analysis and that they are in agreement with the
3 recommendations .
4 As I was reading it through, I was trying to kind
5 of notice what differences , if any, are still there .
6 And I really actually didn' t notice a whole lot .
7 Could you go over that for me as far as what
8 differences there are there in terms of how you' re
9 seeing this and how. Tosco is seeing it .
10 MR . ALTON : I don' t think there - - I think as far
11 as sequence of events, the actual facts , there are no
12 differences .
13 I think we all agree that something happened down
14 on the, bottom of Bed 4 , we don' t know what it is , as far
15 as triggering it is concerned.
16 We all agree there was confusion on the part of
17 the operators . And I think maybe a basic difference ,
18 and maybe only because we ' re going - - we ' re saying it
19 and Tosco is not saying it possibly is that we feel the
20 operators really should be trained, told that they are
21 expected to, that management wants them to, when faced
22 with this kind of a situation, shut the plant down . I
23 think that ' s our basic difference .
24 SUPERVISOR ROGERS : But Tosco' s position, as I
25 understand, is that they did tell the personnel that
Certified Shorthand Reporters
�iBItdOlaerTTB
70 2321 Stanwell Drive•Concord.CIA 94520-4808
REPORTTNG'SERVICE.INC. P.O.Box alo •Conead.CIA 94524- 10
(510)685-6222•Fax (510)685-3829
1 before , - and they will continue to reinforce that .
• 2 MR.. ALTON : But we don' t know how reinforced that
3 was or not . We don't know in what manner it was given
4 -to ` them.
5 SUPERVISOR ROGERS : In other words , was it in a
6 manual somewhere - -
7 MR . ALTON : Yeah.
8 SUPERVISOR ROGERS : - - in some speech, or was it
9 something where somebody really takes somebody aside and
10 says , . hey, this is really important .
11 MR . ALTON : And Supervisor Gerber' s point- is very
12 well taken . I ' ve been in operation before . Operators
13 really want to do whatever they can do to keep a plant
14 running . And most of the time they do an excellent job
15 doing that
16 SUPERVISOR ROGERS : , Thank you .
17 CHAIRMAN DeSAULNIER : All right . Our first
18 .. speaker is Donald Brown . Mr . Brown will be followed by
19 Tom Lindemuth.. And then Jim Payne .
20 MR . BROWN: Good afternoon Supervisors , Chairman
21 DeSaulnier . Very disappointing day here , and my heart
22 goes out to the Glanzman family.
23 .. The major root cause that I look at this - - we ' ve
24 sent you a lot of paperwork on this - - is that the
25 instrumentation was improper. It was like sending
'^ Certified Shorthand Reporters
Za
r 11dond1& 71 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 410;•Concord.CA 94524-4107
- -- — (510)685-6222•Fax(510)685-3829
1 people to war without the proper weaponry, is the best
• 2 way I can- explain it .
3 You had antique equipment that you gave these
4. operators to operate with. And they did everything they
5 could to keep that plant running. Unfortunately, one o.f
6 them got killed and 46 others got injured..
• 7 As I sit here today, it ' s like if I ' m in a
8 neighborhood and I realize at a street corner that
9 there ' s a problem with traffic , I don' t want to wait
10 until that child gets hit and run over to put that light
11 in . I want it put in ahead of time .
.12 Now, the manager, Mr . Miller, came. up here , and a
13 number of times I .heard state-of-the-art equipment .
14 That wasn' t in place , state-of-the-art equipment .
15 And if you look at the process safety management,
16 which we argued over the last year on this good neighbor
17 ordinance , you would see that the first thing that ' s
18 required in process safety management is good
19 engineering control .
20 Nobody mentioned that these weren' t in place
21 today, and this irks me .
22 Now, I was a hydrocracker operator . I know how
23 these units run . So you know, you can come up here and
24 everything can be sweet , but I ' m not going to be
25 bullshitted . You' re not going to tell me how these
i Certified Shorthand Reporters
-�.�iBItdUlteTiB
72 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTTNG SERVICE.INC P.O.sn.4107•concord,CA 64524-410
(510)685-6222• Fax(510) 685-3829
1 things run . I know how they run .
• 2 And what happened in this particular case is
3 instrumentation was faulty, we waited too long to put it
4 in place .- We can not longer afford that .
5 Now, Tosco, they just recently purchased all the
6 Unocal assets out here , 1 . 6 - - $1 . 8 billion. Now, if
7 you' re going to built a refinery state-of-the-art, it
8 will probably cost you that for one refinery. You have
9 four refineries that you got for pennies on a dollar, so
10 there ' s no reason why you can' t take a portion .of your
.11 profits and put them back into safety equipment .
12 There ' s just no excuse for it .
13 And I also realize that there' s been. a number of
14 other incidents happening, and I ' ve looked at those
15 investigations on those . Again, you report that it ' s
16 the operators ' fault .
17 Well , I look at it and I see there ' s a lot of
18 equipment failure there where you failed to recognize
19 that there was a problem. You failed to listen to the
20 operators . You failed to make the improvement .
21 Therefore , it not only cost you to shut down, but it
22 also includes the community having to suck these
23 releases as they come out in the place .
24 You' re still using the same mechanism you used
25 before , and it needs to change . You can come up here
Certified Shorthand Reporters
JrZandondia 73 2321 Stanwell Dri%e•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Bos 4107• Concord.CA 94524-4107
(510) 685-6222 0 Fax(.510) 685-3829
1 and speak all the rhetoric you wish, but in fact it ' s
2 bullshit and it needs to be corrected .
3 We can no longer afford to have these accidents
4 happen on a recurring basis such as they have . And
5 improvements need to be made immediately-
6 And I ' d like to thank you for the opportunity to
7 speak today.
8 . CHAIRMAN DeSAULNIER : Next speaker is Tom
9 Lindemuth . And Mr . Lindemuth will be followed by Jim
10 Payne .
11 SUPERVISOR ROGERS : Mark, could I -
12 CHAIRMAN DeSAULNIER : Supervisor Rogers .
13 SUPERVISOR ROGERS : Could I ask staff a question
14 while - - while the speaker is coming u .
P g P
15 Could staff respond to the comment which - - if I
16 can summarize it a little bit , is that the speaker did
17 not believe that best available technology was in place
18 prior to the incident .
19 I understand that Tosco is committed now to
20 putting in best available technology, but the speaker
21 believed that it was not in place prior to that . Can
22 you comment on that - - on that comment .
23 MR. ALTON : Am I on or off? I ' m on .
24 CHAIRMAN DeSAULNIER : You' re on .
25 MR. ALTON : Okay. Is it the best available? The
Certified Shorthand Reporters
�Lxi��Olte�l8 '� 74 2321 Stamvell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Bos 4107•Concord,CA 94524-4107
RE (510) 685-6222•Fax(510)685-3829
1 instrumentation they were using was fairly modern
• 2 instrumentation . Tosco has upgraded the instrumentation
3 in that plant every few years since it was started up .
4 And they're revolving into the most modern stuff , which
5 you have right now, the so-called Foxboro
6 (unintelligible) distributed control, system, was not
7 functioning at the time .
8 What they were using was -not ancient by any
9 means . I ' m trying to remember, it might be like ten
10 years old or something like that . It ' s not ancient .
11 It had given problems in the past , and that ' s why
12 the operators distrusted it . But I wouldn' t say they
13 were operating without anything at all to. work with.
• 14 CHAIRMAN DeSAULNIER : Thank you .
15 Mr . Lindemuth
16 MR .. LINDEMUTH : Chairman DeSaulnier, members of
17 the Board. As most of you know, I ' m a member of the
18 Hazardous Materials Commission and have been for six
19 years . I ' ve also been a practicing chemical engineer
20 for 30 years , 20 of that here in California .
21 I have comments in a couple of areas . I think
22 first of all , the authors of the report need to be
23 congratulated for the wealth of data that ' s there and
24 the clear and concise way that it ' s presented .
• 25 I have a couple of concerns , however, and some
Certified Shorthand Reporters
�Zaradomfla 75 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.so.4107•concord.CA 94524-4107
(.510) 685-6222• Fax(510) 685-3829
1 generalized recommendations that go with those concerns .
2 First of all , it should be - - I should state that I am
3 not a hydrocracker expert .
4 However, half' a dozen or so years of my
5 experience were in the area of failure mode analysis, in
6 other words looking at complex problems , including
7 sometimes problems. out of my own area of expertise, and
8 helping to guide in the solution of those problems .
9. Part , of that ability is being able to ask questions that
10 nobody else would ask .
11 On the other hand, I do have a fundamental
12 understanding of hydrocrackers , how they - - how
13 hydrocrackers work and have had a chance to review both
14 of the reports rather thoroughly.
15 The first of my concerns is the so-called hot
16 spots that developed in the fourth stage and the fifth
17 stage of the reactor should not be viewed in my opinion
18 as spontaneous occurrences , but I think quite clearly
19 are the result of one to three things or maybe a
20 combination of those things : The design of the system,
21 the operating conditions that it was subjected to in
22 general , or even perhaps more of concern, the kinds of
23 changes and responsive actions that .the operators might
24 have been taking trying to stabilize the unit shortly
• 25 before , let ' s say the day before the event .
�r,,,� i■ Certified Shorthand Reporters
76 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. I P.O.Box 4107•Concord.CA 94524-4107
(510) 685-6222•Fax(510)685-3829
1 I think it ' s clear to everyone that the result of
2 the event, the rapid repressurization of the unit ,
3 probably wiped away a lot of evidence that might be able
4 to have been - - might have been able to be used to get
5 closer to the root cause .
6 But I think it ' s possible and I would urge Tosco
7 to continue to dig in to the true cause of why those hot
8 spots formed when they did, and if wherever possible, to
9 build into their operating procedures commands that
10 would direct the operators -in a - - in a direction that
it would move the - - move the system back into a safe
12 condition .
13 Secondly, the early reliance upon and then
14 rejection of the data from the faulty tempera ture
15 monitoring system illustrates a, truly unfortunate event .
16 I would urge that , as the . unit is repaired and
17 refitted, that truly reliable instrumentation be
18 installed that would sense not only the incipient
19 problem, but also instrumentation that could look at the
20 likelihood that a hot spot would be developing . This is
21 kind of like a stall indicator on an aircraft that
22 combines speed and temperature and so on and says you ' re
23 going to stall .
24 I think it ' s possible with current technology to
25 put together an expert system that may be able to help
Certified Shorthand Reporters
94
77 7 7 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4101•Concord.CA 9$524-4107
(510) 685-6222•Fax(510)685-3829
1 the operators stay out of that - - that, dangerous zone .
• 2 And I think it would be something that would be of
3 benefit , not only to the - - -to the community, but also
4 to Tosco probably from a financial point of view.
5 CHAIRMAN DeSAULNIER: Thank you, Mr . Lindemuth.
6 MR. LINDEMUTH : Any questions?
7 CHAIRMAN DeSAULNIER: No . Thank you .
8 Mr . Payne .
9 MR . PAYNE : Good afternoon, Supervisors . I ' m Jim ,
10 Payne with the. 00AW.
11 I would also like to commend both Tosco and the
12 County for the excellent work they' ve done . I think
13 Tosco assembled a top-notch team of professionals to
• 14 investigate this particular incident . I think they in
15 fact . did a thorough job of investigation, and I would -
16 I think they deserve a lot of credit for what they did.
17 Unfortunately, they let the lawyers write the
18 summary . And I think that the summary that you' ve got
19 before you doesn' t adequately address what their actual
20 results and findings were .
21 If you look at the report , you would think that
22 perhaps the OCAW health and safety representative was an
23 integral part of their investigative team. He in fact
24 was isolated from parts of their investigation . _ And the
25 final report is . not consistent with his beliefs of the
�►^ Certified Shorthand Reporters
��sBILdUItPiTTB
78 2321 Stanwell Drive•Concord,CA 94520-4808
REt EPOMNG SEMCE.INC. P.O.Box 4107•Concord.CA 94524-4107
(510)685-6222-
Fax(510)685-3829
1 way it should have been finalized .
• 2 As a result of. that , he will be putting together
3 a supplemental report for you, giving you his concept or
4 his feelings of what the findings and facts were with
5 the parts of the investigation that he was part of .
6 As it relates to the County' s report , I . think
7 they did an excellent job of going after, trying to
8 assess what the root cause .was and not concentrating on
9 who to blame . I think that ' s the mentality that
10 industry has. to get into .
11 I will also give Tosco credit even for this
12 report being head and shoulders above what anybody else
13 has ever done before, and they do deserve credit for
14 that .
15 As it relates to the County' s report , as I said,
16 I agree with everything that they' ve concluded.
17 However, as far as a number of the management system
. 18 failures , I don' t think they went far enough in
19 addressing those and some of the fixes .
20 Thank you very much.
21 CHAIRMAN DeSAULNIER : Thanks , Jim.
22 SUPERVISOR GERBER : Excuse me .
23 CHAIRMAN DeSAULNIER : Jim, we have a question
24 from Supervisor Gerber .
25 SUPERVISOR GERBER : Excuse me . In your
� ndon,, rr Certified Shorthand Reporters
M
e la 79 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTItVG SERVICE.INC. P.O.sox 410;•Concord.CA 94524-4107
(510)685-6222•Fax (510)685-3829
1 experience , you know, we talk a little bit about what
• 2 kinds of instructions needs to be made to the operators
3 so that they know they have permission which - - you
4 know, bona fide management sanction to depressurize or
5 whatever is called for in a situation similar to what
6 we ' ve been talking about .
7 And I ' m curious as to your opinion about what
8 that_ kind of permission would look like to your
9 coworkers so that they would understand that , in fact ,
10 they' re expected, when in doubt , depressurize .
11 MR . PAYNE : I think it clearly has to be spelled
12 out exactly as you - - as you stated. One of the things
13 for sure that we will do is take a transcript of today's
• 14 proceedings- and let people know that that ' s what - - what
15 management at this time is - - is saying .
16 But I think it ' s absolutely imperative that that
17 be made crystal clear .
18 SUPERVISOR GERBER : Thank you .
19 SUPERVISOR ROGERS : And - -
20 CHAIRMAN DeSAULNIER : Supervisor Rogers has a
21 question, Jim.
22 SUPERVISOR ROGERS : And Jim, from your - - I know
23 that you' re - - you are somebody who works with and talks
24 with a lot of the people who do this kind of work at
• 25 Tosco .
Certified Shorthand Reporters
��iBItdUltPillB
80 2321 Stanwell Drive•Concord,CA 94520-4806
I REPORTING SERVICE.INC P.O.Roz 4167•Concord.CA 94524-4107
j (510)685-6222 0 Fax(510)685-3829
1 What is your feeling about what the perception
2 was amongst those workers . prior to this explosion as far
3 as whether they really did feel comfortable . that if
4 there ' s a close call or a question, that you depressure,
5 or in some other context that you take the action to
6 shut down or slow down production in order to present - -
7 prevent a safety problem?
8 Do you feel that they did have that message prior
9 to this explosion happening?
10 " MR. PAYNE : No, I don' t , and for a number of
11. reasons.
12 One , based on .what Rhear from the folks , that if
13 they depressured every time that happened you' d be
•
14 having - - that incident happened, you' d be having
15 depressuring events happening quite frequently.
16 SUPERVISOR ROGERS : Quite - - .once a week, once a
17 year? I ' m sorry. I don' t
18 MR. PAYNE : I can' t give you a sPecific number .
19 SUPERVISOR ROGERS : Yeah.
20 MR . PAYNE : One operator, and I ' ll give you the
21 number he gave as a - - as a 15-year operator, said that
22 he had seen at least a hundred excursions. that would
23 have caused depressuring, that kind of - - on that kind
24 of a scope
• 25 SUPERVISOR ROGERS : Okay. Thanks .
Certified Shorthand Reporters
(,jZandormI 81 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTItVG SERVICE.INC. P.O.Box 410 •Concord,CA 94524-4107
(510)685-6222•Fax(510)685-3829
1 CHAIRMAN DeSAULNIER : All right . We have another
2 request to speak . Anne Bouguennec - - someday I ' ll
3 pronounce it correctly, Anne . And then we have one
4 other card after that .
5 Given I ' m French, I don"t know why I can' t
6 pronounce a French name .
7 MS . BOUGUENNEC : Good afternoon . You' re getting
8 better at my last name .
9 CHAIRMAN DeSAULNIER: You feel free to
10 mispronounce mine .
11 MS . BOUGUENNEC : Beg your pardon?
12 CHAIRMAN DeSAULNIER : I said feel free to
13 mispronounce mine .
• 14 MS . BOUGUENNEC : Okay. First off , I ' ll take
15 exception to Mr . Brown' s language . I don' t believe it
16 belongs - - that kind of language belongs in a forum.
17 Also the computers at the hydrocracker has only
18 been in service for about four or five years . My
19 husband is a shift supervisor at Tosco, and he has
20 always said and told me - - and he ' s sitting back there
21 now very sort of incensed because he has always told the
22 operators if in doubt , shut the unit down, it isn' t - -
23 let it go and we ' ll work on it later . It ' s always if
24 you' re in doubt , shut it down .
• 25 And that ' s all I ' ve got to say. Thank. you.
�Zandone.IIa Certified Shorthand 520-4Repor8r8
� $2 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SERVICE.I1Vc. P.O.Box 4107•Concord.CA 94524-4107
` --------- --, -- .IT (5101 685-6222•Fax (510)685-3829
1 CHAIRMAN DeSAULNIER: All right_. I have another
2 request to speak . And as near as I can make this out ,
3 it ' s from Mr . Freeman, is that correct? And he would
4 like to speak . on the subject of emergency response .
5 . MR . FREEMAN: Yes . Last week I was at the EPA
6 office for eight hours consecutive in meetings and the
7 review. And the EPA gave me some fact sheets pertaining
8 to. an emergency planning and community right to know
9 act .
10 And I ' d like to ask a question at this time with
11 the Health Department , Mr . Walker, is does he have these
12 fact sheets and is he aware about the EPA regulations
13 (unintelligible) .
�. 14 CHAIRMAN DeSAULNIER : Mr . Freeman - - is that
15 right?
16 MR . FREEMAN : Exactly..
17 CHAIRMAN DeSAULNIER : Is that your name?
18 MR . FREEMAN : Yes , it is , because the two reasons
19 why I ' m asking this question -
20 CHAIRMAN DeSAULNIER : Excuse me .
21 MR . FREEMAN: - - is because I ' d like to know one
22 thing - -
23 CHAIRMAN DeSAULNIER : If I could get your
24 attention for a second .
25 MR . FREEMAN: What?
Certified Shorthand Reporters
�Lxi��� ill8 I'-1
83 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Bos 410 7•Concord.CA 94524-4107
(510)685-6222• Fax(510) 685-3829
1 CHAIRMAN DeSAULNIER: If you' ve got a series of
• 2 questions , you can direct them through the Chair .
3 MR. FREEMAN: Okay. Sure .
4 CHAIRMAN DeSAULNIER: And then we ' ll give
5 Dr. Walker a chance to respond.
6 MR. FREEMAN: Sure .
7 CHAIRMAN DeSAULNIER: ,Thank you .
8 MR. FREEMAN: The reason why I ' m bringing this
9 issue up at this time, because since 1994. catacarb
10 skill , which I was affected in that spill myself as a
11 victim, the victims are the last ones to be notified of
12 any kind of emergency planning and community right to
13 know acts .
• 14 And I ' d like to see at this time that the injured
15 parties of this Tosco incident are very informed with
16 the health - -high health risk evaluation assessments to
1.7 their - = to the injuries of this by the Health
18 Department .
19 And I ' d like to see the Health Department play a
20 more active role in submitting more accurate, up-to-date
21 in SDS reports that could be (unintelligible) to the
22 Board of Supervisors and to other County officials as
23 well .
24 Okay. That ' s my statement . -
25 CHAIRMAN DeSAULNIER : Thank you, sir .
Certified Shorthand Reporters
84 2321 Stamvell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4107•Concord.CA 94524-4107
(510)685-6222•Fax(510)685-3829
1 MR . FREEMAN : Thank you .
• 2 CHAIRMAN DeSAULNIER : That ' s the last speaker
3 card we have , so we will now close the public comment
4 period.
5 Dr . Walker, do you have a response to
6 . Mr . Freeman?
7 DR . WALKER : No .
8 CHAIRMAN DeSAULNIER: As near as I can tell , it
9 was EPA guidelines and - - no?
10 DR . WALKER : Right . No, sir, I don' t have any
11 comment .
12 CHAIRMAN DeSAULNIER : Maybe I ' ll make a
13 suggestion, Mr . Freeman, if you could - - if you could
• 14 put your request in writing and give it to me , I will
15. respond to it , and _ I ' ll ask County staff to help me with
16 the response .
17 So bring it back to the Board for any further
18 discussion, decision . And if I could just lead- off with
19 a couple of comments since it ' s in my district .
.20 I would like to commend staff for the work that
21 you have done . And to Tosco, I ' d like to acknowledge
22 that Mr . Wiggins came out here, the president .of Tosco,
23 and the extensive amount of time that Tosco put in .
24 And also, as I mentioned, the change at least
25 from just a few months ago when we were at lock jams in
Certified Shorthand Reporters
85 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 410 •concord.CA 99524-410
(510) 685-6222•Fax(5 10)688-3829
1 terms of sharing information and .knowing. that you put
• 2 yourself at some risk as a - as a corporation to do
3 this . But I think this is really worthwhile and it ' s a
4 step -down the road.
5 And granted, to Mr . Brown and others , it ' s not
6 far enough, but I think you would have to admit that
7 we ' ve come a long way in the last few years .
8 And maybe what I would say also to the operators
9 who have been here , I know this has been a very
10 traumatic experience , and some of the work that we' ve
11 put into this has caused some stress and some pain . So
12 for me , I would like to apologize for that , but I think
13 that there is - - it ' s a good investment in the long run .
• 14 And just for matters of discussion of the Board,
15 I would propose to make a motion to follow staff ' s
16 recommendations , but also ask for Tosco perhaps to come
17 back before the July startup and just give us an update
. 18 on the 12 bullets and how much of that work has been
19 accomplished . I will take Mr. Miller at his word that
20 those will be completed.
21 And maybe if we could have. a report back on that
22 before startup, and then offer to both CBE and to OCAW,
23 for them to request a time when they would like to come
24 back with followup comments and report , they can do that
25 through the Chair and we ' ll allot time on the schedule
nr_nd0Re*r$ ve*Con ShoConcord,
CA 9 Reporters
Off, 1` � $6 2321 Stan��ell Drive•Concord,CA 94520-4808
REPORTING SERVICE.INC. P.O.Box 4107•Concord.CA 94524-4107
(510) 685-6222•Fax(510)685-3829
-----
1 for that , so - -
2 SUPERVISOR UILKEMA: I ' ll second the motion to - -
3 CHAIRMAN DeSAULNIER: All right . Then I will
4 make it a motion .
5 Any further discussion?
6 SUPERVISOR UILKEMA: Yes . One of the things that
7 I was very interested in were the remarks. by
8 Mr . Lindemuth. And- he was discussing something even
9 more fundamental which we have assumed in the reports as
10 somehow the hot spot occurred. And it seems to be - -
11 it ' s kind of a mystery area .
12 And I would like to pursue that a little more
13 inasmuch as it seems it may not be answerable , but it
14 might be ver helpful to search data in other .refiner
g Y P Y
15. -installations to see whether there is some kind of
16 consistency and. that - - that aspect is something that
17 perhaps we might. want to pursue :
18 I would like to say that I was very pleased with
19 the report by the Health Department , thank Tosco for
20 your reports and for your cooperation, for Mr . Wiggins '
21 trip out here . And Mr . Miller, I think you did an
22 excellent job .
23 And I noticed that one of the things that the
24 community really does appreciate is forthright honesty,
25 and we all make mistakes , and the .willingness to say
�M�r� Certified Shorthand Reporters
ndondhi 87 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTIl`IG SERVICE.INC. P.O.Box 4107•Concord.CA 94524-4107
(510)685-6222•Fax(510)685-3829
1 that you do intend to change . And I believe that ' s
• 2 respected. a great deal by the community. And I ' m very
3 pleased to hear it .
4 And I look forward to things getting better and
5 the level of confidence between the industrial community
6 and certainly the County government to continuously
7 - improve . So I think we' re on the right road.
8 So I would add as a friendly amendment or
9 suggestion that Mr . Lindemuth' s remarks about the issue
10 of the hot spots be included in either remarks for the
1.1 future or the possibility of some kind of further
12 investigation .
13 CHAIRMAN DeSAULNIER : I think that would be an
subject f
14 appropriate subs r the Hazardous Materials o
15 Commission to look at as well , and I ' m sure both of you
16 will bring that up .
17 Supervisor _ Rogers .
18 SUPERVISOR ROGERS : I was after Donna .
19 CHAIRMAN ,DeSAULNIER: Donna? Joe?
20 SUPERVISOR ROGERS : I ' d like to ask with the
21 Board' s indulgence if we could get a brief response from.
22 Tosco . There was a comment by Jim Payne earlier to the
23 effect that there were some difficulties in including
24 the OCAW representative in the Tosco process .
25 If I could ask the Board' s indulgence, if we
� ,, ? Certified Shorthand Reporters Mndol6Il8 88 2321 Stanwell Drive•Concord,CA 94520-4808
REPOI2TIP1Ci SERVICE.INC.
P.O.Box 4107•Concord.CA 94524-4107
(510)68.5-6222•Fax (510)685-3829
1 could get a brief clarification on that or response or ,
• 2 whatever from Tosco, because I ' m concerned by that
3 allegation .
4 CHAIRMAN DeSAULNIER : Would someone in Tosco like
5 to make a comment? Chuck .
6 MR. WAITMAN My name is Chuck Waitman, as I
7 think you know, and I headed the investigation .
8 We did have a multi-disciplined team meet daily.
9 during our investigation for the first three to three
10 and a half months of the investigation . We both
11 compiled and analyzed facts during that period. We also
12 organized our. facts to some degree .
13 The last month of the process was really worked
• 14 in trying to get to a good, sound and thorough public
15 document . And the team did review those documents , but
16 I think we - - obviously John has some things in the
17 final document that. he wants to review with me . And ,I
18 will be more than pleased to do so, and also more than
19 pleased to have him come back and present any lack of
20 agreement he has to you next month.
21 So I think we felt we were inclusive , but
22 obviously in the final draft there ' s some level of
23 disagreement. And we' re all for sorting out those
24 disagreements .
25 Thank you .,
Certified Shorthand Reporters
��,BIadUItP,TTB
89 2321 Stanwell Drive•.Concord,CA 94520-4808
REPORTING SERVICE INC P.O.Box 4107•Concord.CA 94524-4107
(510)685-6222•Fax(510) 685-3829
l CHAIRMAN DeSAULNIER : And I think we ' re probably
2 hear - -obviously hear more of this when OCAW comes back
3 to us .
4 SUPERVISOR ROGERS : I guess ,. I wasn' t so much
5 concerned about the fact that he had a different
6 opinion, and I think that ' s - - that ' s fine
7 I guess what I was asking is , if I understood
8 Mr . Payne right, he was saying he wasn' t complaining.
9 about the fact that there was a difference of opinion .
10 He was complaining about the fact that he didn' t feel
11 that the OCAW person was legitimately at the table
12 during the Tosco committee ' s process .
13 Can you comment on that?
14 MR . WAITMAN : Well , I felt that he was at the
15 table for - - you know, on a daily basis for most of many
16 months , so - -
17 SUPERVISOR ROGERS : Thank you .
18 SUPERVISOR GERBER: Was he at all the meetings of
19 the group?
20 MR . WAITMAN: He was never excluded from a
21 meeting, and he was certainly at most of them. I dont
22 know that any of us were at every meeting for a three
23 and a half month period .
24 CHAIRMAN DeSAULNIER: Jim is shaking his head up
• 25 and down that he was .
q,,,� Certified Shorthand Reporters
L�aItdUlt6lI8
90 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.sox 4107•Concord,CA 94524-4107
(510)685-6222•Fax(510)685-3829
1 MR., WAITMAN: Okay.
• 2 CHAIRMAN DeSAULNIER : All right . Maybe - -
3 MR . PAYNE : Do you want a specific comment from
4 me on that?
5 CHAIRMAN DeSAULNIER: All right .
6 MR . PAYNE : John participated in most of the
7 meetings that took place . When the interviews with
8 management personnel over what they did, what they knew
9 and all that took place , he was specifically excluded
10 from those meetings .
11 CHAIRMAN DeSAULNIER : All right . Thanks, Jim.
1.2 Well , we - will anticipate a report back from OCAW
13 and your representative on what his understanding. was
14 that the final report would include and then some
P include ,
, and
comments about where he was included and where he was
16 not , when it comes back to us .
17 Jim, is that enough?
18 SUPERVISOR ROGERS : Yeah. And I was - - I was
19 going to say I know it ' s a difficult process to do the
20 kind of work that Tosco has done here, and it is a much
21 more factual report than what we ' ve been accustomed to
22 getting, and so I don' t mean to downplay that or
23 minimize that .
24 But at the same time one part of this process is
25 trying to have something out there where the public both
M,,.� Certified Shorthand Reporters
ildo tPiTTB 1'� 91 2321 Stanwell Drive•Concord,CA 94520-4808
REPORnNG SERVICE.IN'Cl P.O.Boz 4107•Concord,CA 94524-4107
(510)685-6222•Fax(510)685-3829
1 feels they can understand what happened and they can
• 2 trust the process that brought the report forward .
3 And I think if there are future efforts like
4 this , I think it is really -important to make sure that
5. OCAW or whatever other workers ' representatives are
6 really in the loop as much as possible .
7 And I think if that happens , it really benefits
8 everybody including Tosco because it makes the final
9 product much more believable to the public .
10 CHAIRMAN DeSAULNIER : All right . With that , I ' ll
11 call for the question . All those in favor . Aye
12 (Ayes . )
13 CHAIRMAN DeSAULNIER : Opposed? That was - - that
14 was five to nothing, believe it or not .
15 - -000- -
16
17
18
19
20
21
.22
23
24
25
�r� allR.
Certified Shorthand Reporters
�zandolm92 2321 Stanwell Drive•Concord,CA 94520-4808
.REPORTING SERVI ! P.O Bo.410-•concord.C.�94524-4107
(510)685-6222• Fax (510)685-3829
1 STATE OF CALIFORNIA )
ss .
• 2 COUNTY OF CONTRA COSTA )
3 I , JOHN A. ZANDONELLA, do hereby certify:
4 That I. am a Certified Shorthand Reporter of the
5 State of California, License No . C-795 ;
6 That the foregoing pages are a true and correct
7 transcript of the tape-recorded proceedings before the
8 Contra Costa County Board of Supervisors, County
9 Building, Board Chambers , Martinez , California, except
10 as noted "unintelligible" or " inaudible . "
11 I further certify that I am not interested in the
12 outcome of said matter nor connected with or related to
13 any of the parties of said matter or to their respective
• 14 counsel .
15 Dated this 12th day of June, 1997 .
16
17 - -000- -
18
19
20
21
22 JOHN A. ZANDONELLA, CSR License No . C-795
23
24
25
Certified Shorthand Reporters
�ZandUlad —� 93 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. !i
P.O.Box 4107•Concord.CA 94524-4107
_ __� (5]0)680-6222•Fax (510)685-3829
1 terms of sharing information and knowing that you put
• 2 yourself at some risk as a - - as a corporation to do
3 this . , But I think this is really worthwhile and it ' s a
4 step down the road .
5 And granted, to Mr . Brown and others , it ' s not
6. far enough, but I think you would have to admit that
7 we ' ve come a long way in the last few years .
8 And maybe what I would say also to the operators
9 who have been here , I know this has been a very
10 traumatic experience , and some of the work that we ' ve
11 put into this has caused some stress and some pain . So
12 for me , I would like to apologize for that , but I think
13 that there is - - it ' s a good investment in the long run .
14 And just for matters of discussion of the Board,
15 I would propose to make a motion to follow staff 's
16 recommendations , but also ask for Tosco perhaps to come
17 back before the July startup and just give us an update
18 on. the 12 bullets and how much of that work has been
19 accomplished. I will take Mr . Miller at his word that
20 those will be completed.
21 And maybe if we could have a report back on that
22 before startup, and then offer to both CBE and to OCAW,
23 for them to request a time when they would like to come
24 back with followup comments and report , they can do that
• 25 through the Chair and we ' ll allot time on the schedule
�fi�rr,,,,.,� � Certified Shorthand Reporters M]RdoxmHa 86 2321 Stanwell Drive•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Boz 410 •Concord,CA 94524-4107
— (510) 685-6222•Fax(510) 685-3829
1 for that , so - -
• 2 SUPERVISOR UILKEMA: I ' ll second the motion to - -
3 CHAIRMAN DeSAULNIER : All right . Then I will
4 make it a motion.
5 Any further discussion?
6 SUPERVISOR UILKEMA: Yes . One of the things that
7 I was very interested in were the remarks by
8 Mr . Lindemuth. And he was discussing something even
9 more fundamental which , we have assumed in the reports as
10 somehow the hot spot occurred. And it seems to be , - -
11 it ' s kind of a mystery area . .
12 And I would like to pursue that a little more
13 inasmuch as it seems - - it may not be answerable , but it
• 14 might be very helpful to search data in other refinery
15 installations to see whether there is some kind of
16 consistency and that - -that aspect is something that
17 perhaps we might want to pursue .
18 I would like to say that I was very pleased with
19 the report by the Health Department , thank Tosco for
20 your reports and for your cooperation, for Mr . Wiggins '
21 trip out here . And Mr . Miller, I think you did an
22 excellent job .
23 And I noticed that one of the things that the
24 community really does appreciate is forthright honesty,
25 and we all make mistakes , and the willingness to say
Certified Shorthand Reporters
�ZaI�SEC.
87 2321 Stanwell Drive•Concord,CA 94520-4808
REPORT P.O.Box 4107•Concord.CA 94524-4107
(510)683-6222•Fax(5,10)685-3829
1 that you do intend to change . And I believe that ' s
• 2 respected a great deal by the community. And I ' m very
3 pleased to hear it .
4 And I look forward to things getting better and
5 the level of confidence between the industrial community
6 and certainly the County government to continuously
7 improve . So I think we ' re on the right road.
8 So I would add as a friendly amendment or
9 suggestion that Mr . Lindemuth' s remarks about the issue
10 of the hot spots be included in either remarks for the
11 future or the possibility of some kind of further
12 investigation .
13 CHAIRMAN DeSAULNIER. I think that would be an
14 appropriate subject for the Hazardous Materials
15 Commission to look at as well , and I ' m sure both of you
16 will bring that up .
17 Supervisor Rogers .
18 SUPERVISOR ROGERS : I was after Donna
19 CHAIRMAN DeSAULNIER : Donna? Joe?
20 SUPERVISOR ROGERS : I ' d like to ask with the
21 Board' s indulgence if we - could get a brief response from
22 Tosco . There was a comment by Jim Payne earlier to the
23 effect that there were some difficulties in including
24 the OCAW representative in the Tosco process .
25 If I could ask the Board' s indulgence , if we
i Certified Shorthand Reporters
�ZBItdUItPiIIB $$ 2321 Stam+ell Dri%e•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O. Box 4107•Concord,cA 94524-4107
- (510)68.5-6222•Fax(510)685-3829
1 could get a brief clarification on that or response or
2 whatever from Tosco, because I ' m concerned by that
3 allegation .
4 CHAIRMAN DeSAULNIER: Would someone in Tosco like
5 to make a comment? Chuck .
6 MR . WAITMAN: My name is Chuck Waitman, as I
7 think you know, and I headed the investigation .
8 We did have a multi-disciplined team meet daily
9 during our investigation for the first three to three
10 and a half months of the investigation . We both
11 compiled and analyzed facts during that period . We also
12 organized our facts to some degree .
13 The last month of the process was really worked
14 in tryingto get to a good, sound and thorough public
15 document . And the team did review those documents , but
16 I think we - - obviously John has some things in the
17 final document that he wants to review with me . And I
18 will be more than pleased to do so, and also more than
19 pleased to have him come back and present any lack of
20 agreement he has to you next month.
21 So I think we felt we were inclusive, but
22 obviously in the final draft there ' s some level of
23 " disagreement . And we ' re all for sorting out those.
24 disagreements .,
25 Thank you .
Certified Shorthand Reporters
r �Zandonei 11 89 2321 Stanwell Drive•Concord,CA 94520-4808
REPORTING SEMCE.INC. P.O.Box 410;•Concord.CA 94524-4107 .
_� (510)685-6222•Fax(510)685-3829
1 CHAIRMAN DeSAULNIER: And .I think we' re probably
2 hear - -obviously hear more of this when OCAW comes back
3 to us .
4 SUPERVISOR ROGERS : I guess , I wasn' t so much
5 concerned about the fact that he had a different
6 opinion, and I think that ' s that ' s fine .
7 I guess what I was asking is , if I understood
8. Mr . Payne right , he was saying he wasn' t complaining
9 about the fact that there was a difference of opinion .
10 He was complaining about the fact that he didn' t feel
11 that the OCAW person was legitimately at the table
12 during the Tosco committee ' s process .
13 Can you comment on that?
• 14 MR. WAITMAN : Well , I felt that he was at the
15 table for - - you know, on a daily basis for most of many
16 months , so - -
17 SUPERVISOR ROGERS : Thank you .
18 SUPERVISOR GERBER : Was he at all the meetings of
19 the group?
20 MR . WAITMAN : He was never excluded from a
21 meeting, and he was certainly at most of them. I don' t
22 know that any of us were at every meeting for a three
23 and a half month period .
24 CHAIRMAN DeSAULNIER : Jim is shaking his head up
25 and down that he was .
Certified Shorthand Reporters
�— ZandUnefla 90 2321 Stanwell Drive•Concord.CA 94520-4808
i REPORTING SERVICE.INC. P.O.Bo.410 •concord.CA94524-410-
(510)683-6222• Fax(510)685-3829
1 MR. WAITMAN: Okay.
2 CHAIRMAN DeSAULNIER : All right . Maybe - -
3 MR . PAYNE : Do you want . a specific comment from
4 me on that?
5 CHAIRMAN DeSAULNIER : All right .
6 MR. PAYNE : John participated in most of the
7 meetings that took place . When the interviews with
8 management personnel over what they did, what they knew
9 and all that took place , he was specifically excluded
10 from those meetings .
11 CHAIRMAN DeSAULNIER : All right . Thanks , Jima
12 Well , we will anticipate a .report back from OCAW
13 and your representative on what his understanding was
14 that the final report would include , and then some
15 comments about where he was included and where he was
16 not , when it comes back to us .
17 Jim, is that enough?
18 SUPERVISOR .ROGERS : Yeah . And I was - - I was
19 going to say I know it ' s a difficult process to do the
20 kind of work that Tosco has done here , and it is a much
21 more factual report than what we ' ve been accustomed to
22 getting, and so I don' t mean to downplay that or
23 minimize that .
24 But at the same time one part of this process is
25 trying to have something out there where the public both
* Certified Shorthand Reporters
ndUItPTTB 91 2321 Stanwell brise•Concord.CA 94520-4808
REPORTING SERVICE.INC. P.O.Bo.416 •Concord.CA 94524-1107
(510) 685-6222•Fax(510)685-3829
l feeds they can understand what happened and they can
2 trust the process that brought the report forward .
3 And I think if there are future efforts like
4 this, " I think it is really important to. make sure that
5 OCAW or whatever other workers ' representatives are
6 really in the loop as much as possible .
7 And I think if that happens , it really benefits
8 " everybody including Tosco because it makes the final
9 product much more believable to the public .
10 CHAIRMAN DeSAULNIER : All right . With that , I ' ll
11 call for the question . All those in favor . Aye .
12 (Ayes . )
13 CHAIRMAN DeSAULNIER : Opposed? That was - = that
14 was five to nothing, believe it or not .
15 - -000- -
16
17
18
19
20
21
22
23
24
25
1O�ei*� Certified Shorthand Reporters
1
(j 1 92 2321 Stanwell Drive•Concord,CA 94520-4805
REPORTING SEMCE.INC. P.O.Box 410 •Concord,C9 94524-410
— – _ -- (510)685-6222•Fax(510) 685-3829
c
1 STATE OF CALIFORNIA )
ss
2 COUNTY OF CONTRA COSTA )
3 I , JOHN A. ZANDONELLA, do hereby certify:
4 That .I am a Certified Shorthand Reporter of the
5 State of California, License No . C-795 •
6 That the foregoing pages area true and correct
7 transcript of the tape-recorded proceedings before the
8 Contra Costa County Board of Supervisors , County
9 Building, Board Chambers , Martinez , California, except
10 as noted "unintelligible" or " inaudible . "
11 I further certify that I am not interested in the
12 outcome of said matter nor connected with or related to
13 any of the parties of .said matter or to their respective
14 counsel .
15 Dated this 12th day of June , 1997 .
16
17 - -000- -
18
19
20
21
22 OHN A. ZANDONELLA, CSR License No, C-795
23
24
25
Certified Shorthand Reporters .
�Zandone la 93 2321 Stanwell Drive a Concord.CA 94520-4808
REPORTING SERVICE.INC. i P.O.sow 1107•concord.CA 94524-1107
(510)685-6222•Fax(510)685-3829.