LG Polymers Accident in AP: Critique of the Joint Monitoring Committee Report

June 6, 2020

 This critique was received from Scientists for People 

Joint monitoring committee appointed by NGT, Principal Bench, New Delhi submitted a report on 28 May 2020. We could get a copy thanks to a journalist on 30 May 2020. After going through the report we are disappointed with the scientific quality and integrity of the report.

Report has contents that are a verbatim reproduction from different sources without acknowledgement as references. In a less than 20 page written part of the report, the committee of experts resorted to plagiarism and we consider it as ethical misconduct.

In “Critical Aspects of Safety and Loss Prevention” Trevor A Kletz suggests 

“An accident report should tell us: 1. What happened; 2. Why it happened 3. What we should do differently in future to prevent it happening again, on other plants as well as on the plant where it actually occurred. 4. Who should make the changes? 5. When the changes will be complete. The report can then be brought forward at this time. 6. What the changes will cost.” 

Though the purpose of the report is somewhat different, it should meet these and similar requirements to meet the purpose. 

We first present the copy pasted parts with sources unacknowledged and then critique the investigative work supposedly done. 

Plagiarism: 

Entire text reproduced here was copy pasted into the report from the web resources listed below the text. 

(1) 5 Properties of Styrene (Page 11) : “Styrene is a colorless, clear liquid. It has a sweet smell and can be found in nature as well as manufactured. Styrene was originally found in the oriental sweet gum tree (levant styrax). It can also be found in common foods and beverages, such as strawberries, coffee, cinnamon, peanuts, and tobacco. Manufactured styrene has a wide range of uses and is a component of many goods, including: polystyrene, fiberglass, packaging materials, electrical insulation, home insulation, drinking cups and food packaging, rubber, and carpet backing. “

Source of plagiarism:

https://u.osu.edu/helmig-mason.1/tag/styrene/2

(2) “Styrene is a volatile, highly flammable compound. Styrene vapour is heavier than air. At concentrations normally encountered in the workplace, the air and styrene mixture is not significantly heavier than clean air. Styrene evaporates more rapidly at high temperatures (e.g. the evaporation rate at 30° C is twice that at 20 ° C). Styrene can be smelled at very low concentrations. Prolonged exposure to styrene reduces a person’s ability to smell it. Styrene liquid is soluble in body fat and can be absorbed through the skin, however, studies have shown that the styrene present in polyester resin is not easily absorbed through the skin. Inhalation is therefore the major route of exposure. “

Source of plagiarism:

https://www.commerce.wa.gov.au/sites/default/files/atoms/files/code_styrene.pdf

(3) Page 15, Para 3: “It should be noted that in climate zones and in seasons with significant temperature differences between night and day, the styrene vapours evolved in the headspace at higher temperatures will condense on roofs, walls and internal fittings of storage tanks when it cools off. The phenolic inhibitors have high boiling points and stay in the liquid phase, resulting in the condensed styrene vapours containing no inhibitor.” 

Source of plagiarism: Copied from Page 12 of Styrene Monomer: Safe Handling Guide, July 2018 

https://www.plasticseurope.org/download_file/view/2150

(4) “Experiments showed that the polymerization runaway “onset” temperature inversely increased with the monomer mass fraction. Styrene polymerization reaction is relatively highly exothermic with a heat generation at around 71 kJ·mol−1. At the same time, even without an initiator, two styrene molecules will undergo a Diels−Alder type of reaction and generate radicals to start self-polymerization upon heating “

These sentences above were verbatim copied into the report on page 15 from the publication referred below. 

[Lin Zhao, Wen Zhu, Maria I. Papadaki, M. Sam Mannan, and Mustafa Akbulut, Probing into Styrene Polymerization Runaway Hazards: Effects of the Monomer Mass Fraction, ACS Omega 2019, 4, 8136−8145] 

Wasted opportunity: 

With heavy heart, we are forced to state that the Joint Monitoring Committee appointed by NGT has wasted an opportunity to investigate the release of styrene incident. NGT has erred in assembling a team of academics without any exposure to industry or process safety. The report is a futile exercise devoid of any actual investigation.

It is a mere compilation of the information given by the company. Committee allowed itself to become a proxy to the company by uncritically accepting every information fed to it. LG Polymers successfully strategized their stand to prevent the real truth revealed in the investigation.

Expert team miserably failed to expose LG Polymers for data fudging. Entire report has no field data for the styrene storage system. Even a simple flow sheet for the tank is not given. Sketch of the tank with all its nozzles, roof vents and measuring points is not presented. 

Many accidents are reported during startup of chemical plants. Management pressure pushing to rush meet the schedules have led to accidents. Was there any such target due to plant shutdown during the lockdown period? Investigators did not consider the startup process for any role in the accident. 

Report does not indicate any real effort by the committee to rebuild the scenario at the time of the accident. We wonder how some of the committee members who never visited the plant approved the report. One of the important ethical codes for persons in science and engineering is “Perform professional services only in areas of their competence” and “Hold paramount the safety, health and welfare of the public and protect the environment in performance of their professional duties.”

The committee has failed in both respects through their inadequacies. It appears that the committee did not investigate to get at the truth but allowed itself to justify preconceived opinions on the accident. 

Shortcomings of the report: 

Presented below are quick comments obtained from Sri B Sri Ramachandra Sai, who was a senior general manager and head of department Process and Mechanical at Thyssenkrupp industrial systems India Pvt. Ltd. Mumbai, a major engineering corporation specializing in project design and engineering of large chemical industries like fertilizers and petrochemicals. He has experience of designing and engineering styrene production plants. 

1. The report did not cover any aspects regarding the international practices in designing, operating and maintaining styrene storage installations and comparing what is at present at LG Polymers. 

2. No operating data regarding how the storage system is operated from the day of lockdown till the accident for all storage tanks of Styrene at plant as well as onshore tanks at port. DCS graphics operating data is missing. 

3. No mention of plant SOPs regarding prevention of runaway condition for example, addition of diluents to stop polymerization etc. 

4. No mention of the condition of other styrene storages at the time of accident 

5. There is no mention of sampling frequency and analysis of styrene during lockdown period for polymer content to use as guidance to take steps to prevent auto polymerisation. Temperature measurement cannot be the only guidance factor to the operator to prevent auto polymerisation. 

6. No mention of HAZOP report required to be generated as recommended practice whenever plant is installed and whenever modifications are carried out in the plant 

7. No mention of dispersion study done for the case of uncontrollable runaway condition occurring in the largest storage in the premises.

8. No mention of rate of discharge when the accident happened and the pressure build up in the tank 

9. No mention of deficiencies in design to ensure uniform temperature inside the tank and when the calibration is done for temperature sensor on the tank 

10.No mention is made regarding design deficiencies to ensure 15 ppm of inhibitor TBC in the storage 

11.No mention of any data regarding cooling system and control to ensure less than 20 deg.C in the tank 

12.No mention of gas detection system and deluge system available in the plant 

13.No mention of temperature gradient from top to bottom observed by plant personnel during so many years of operating the styrene storage tanks during peak summer to guide them to take preventive steps leading to auto polymerisation 

14.No as built P&I diagram of storage is attached. Also no Tank design drawing with nozzle schedule is attached 

15.No mention of layout of storage system with respect to plant boundary and to the nearest habitat meeting statutory requirements 

16.No mention of containment of styrene liquid in case of leakage and operating procedures to prevent spreading to environment 

17.No mention regarding reasons for fast spreading of styrene vapor to nearest village. Any water addition to cool evaporating styrene causing violent boiling of water causing faster dispersion of styrene vapor along with water vapor could be the reason? 

18.Gooseneck vent nozzle and Dip hatch nozzles are the nozzles through which styrene vapor is released. But no mention is made regarding safety vent nozzles on the tank 

19.No mention regarding whether polymerization had happened inside circulating line causing failure of circulation 

20.No mention is made regarding emergency systems that are connected to no break power. For example: circulating system, cooling system, inhibitor addition system or emergency diluent addition system etc. 

21.No mention regarding how the tank temperature is to be brought down and what are the future recommendations for faster cooling of tank before vapor is spread to environment 

The above comprehensive comments by an expert clearly bring out invalidity of the present report for basing any decision by the court and the need for a better team to professionally and thoroughly investigate the accident. 

Though Scientists for People have been sending useful information for helping the investigation, such as calculations for estimating the total styrene release into the environment by different methods, styrene vapour cloud dispersion modeling done by us and also offered to come online to interact with them. For undertaking root cause analysis we offered to bring online Sigma expert well versed in root cause analysis procedures. But the expert team never responded. 

Here we present a few glaring errors in the report along with our critique

1. “The sequence of operations carried out at the factory after the incident as per plant managers is as follows:” (Page 12) 

The list given by the company is presented on pages 12 and 13 of the report as it is without any comments/observations of the committee. Uncritically accepting the inputs from the company under investigation negates the very purpose of appointing the committee. 

03:02 hrs: M6 Tank temperature started rising. 

03:07 hrs: Alerted security in-charge to get help from outside agencies (Fire services and Ambulance etc.). Root cause was identified as self-polymerization due to stagnant high polymer content. [emphasis added] 

22:45 hrs: Tank temperature reached 154°C. 

What is the type of temperature measurement used? What were the temperatures like during the period before and during 03:02 hrs to 22:45 hrs on 7 May 2020? Why has the committee not obtained the temperature history inside the tank? Why is that data not presented in the report? Who identified the root cause as self polymerization? In root cause analysis the next question is why did self polymerization happen?

Self polymerization can be one of the causes but not the root cause. It is a misuse of the word ‘root cause’. How did they conduct any analysis in the panic situation and conclude the cause? One of the committee members in an interview to BBC Telugu repeated the company stand that siren was blown but people did not notice it due to panic. It did not occur to the expert how the people became panicky even before they were alerted on styrene release?

With such alertness and analytical capabilities nothing better can be expected from that team. 

Tank temperature reaching 154 C when the normal boiling point is only 145 C needs explanation. But the committee simply ignored this fact. The report has not considered the pressure aspects in the tank. What was the maximum temperature reached in the tank? At 155 C the vapour pressure of styrene is 969.78 mm Hg or 1.27 atm. 

“The pressure in the storage tank will progressively increase, and the safety valves released the styrene vapour into the atmosphere. The increase in temperature and pressure was not observed by the industry.” (Page 14) 

That is a vague statement. Expert team did not care to look into to the abnormal claim of LG Polymers on no increase in temperature and pressure. Why have they ignored investigation of such unusual phenomena? Is there a pressure measurement in the tank? Above statement is valid only when the pressure in the tank is measured regularly.

Has the team accessed any pressure data to verify the industry claim? If the pressure did not really increase, expert team should have presented an explanation for it. In absence of pressure increase the expert team has to explain reasons for boiling point elevation for styrene.6 

The timeline entries are incomplete and contradicting to some observations by the committee. For example, “The leaked tank was old and does not have temperature sensors at middle and top surface of the tank except only provision to measure the temperature at the bottom of the tank where refrigeration is provided.” (Page 16) 

“The leaked tank does not have any provision for measuring the vapour space temperature. Due to this, building-up of temperatures in top surface could not noticed by the industry.” (Page 15) 

Have the experts physically verified the lack of temperature measurement except at the bottom? Why have they not obtained the design drawing of the tank and its P&I for verification? Are there no provisions on the tank for measuring temperature at different heights or the company did not install a temperature sensor in the thermowells provided? Report lacks thoroughness. 

The list refers to DCS system. But the experts have not referred to it in the report or tried to make use of it to call the bluff of the company in feeding them sanitized information. 

There are no internals in styrene storage tanks as a standard practice. We understand that styrene is cooled in an external cooling circuit. Internals are avoided to eliminate surfaces promoting polymerization. Internal surface of the tank is lined to reduce any possibility of surface induced polymerization.

Expert team being the only team that had access to the plant at least on three days, it is clear that their lack of knowledge in process industries prevented them to observe the plant critically. There are no details of cooling, circulation and refrigeration systems. The word refrigeration occurs only two times in the report and there are no details about it and whether it was working or not during the crisis. 

Among the six steps in an accident investigation according to “Guidelines for investigating process safety incidents” “The Third Step in incident investigation is to gather information, separate facts from suppositions, analyze data, and determine what happened. Before conducting a cause analysis, a comprehensive and accurate understanding of what happened must first be completed. Witness management, evidence management, and evidence analysis and hypothesis testing are key concepts to be employed during the investigation process.” 

The expert committee has not presented any data collected in the process of its investigation. “We should include in accident reports all the facts that have come to light, even though no conclusions are drawn from some of them, so that readers with different backgrounds, experience or interests can draw additional conclusions which were not obvious to the original investigators.” Author gives two examples of how readers came out with better solutions to the problem. [Trevor A Kletz, Learning from Accidents, 2001] 

Any conclusions drawn are as good as the quality of the data and absence of data makes the report invalid.

2. “On the early hours of May 07, 2020, the tank with 1830 tons of storage had developed the leak of the styrene vapours from the top of the tank and spread beyond the factory boundary towards the west side due to wind direction and affected the residents of 5 nearby areas namely, Venkatapuram, Venkatadri Nagar, Nandamuri Nagar, Pydimamba Colony and BC colony.” (Page 14) 

Styrene stored at below 20 C does not suddenly begin to boil to release styrene vapours. For 1800 T of styrene present in the tank to reach boiling temperature, 96 million kCals of heat has to be generated within the storage tank. LG Polymers had clear three days after the lifting of lockdown and began preparations for regular operation of the plant before the incident occurred. Staff were into normal start up mode.

There must have been plenty of indications, several of them proxy, about the change of temperature in the tank if only anyone cared to recognize. It was reported in New York Times that the clogged cooling system was responsible for the incident according to three anonymous experts in the committee Reuters spoke to. Is that true? The report is silent on cooling, recirculation and refrigeration systems. Why? Are they unimportant in the investigation?

With the circulation off and cooling system shutdown, even with inhibitor present in styrene, increase in temperature is likely especially in the summer month of May. Expert team has not conducted one – on – one interviews with operating staff present during the incident to reconstruct the developments during the lockdown period and since 4 May onwards. Instead of investigation the team resorted to wishful speculation from preconceived notions. Objectivity is lost. 

3. “Root cause analysis showed that the problem possibly began on April 20, 2020 when the polymer concentration in Tank M6, which was idled at full capacity since March 25 post lockdown.” (Page 15) 

This sentence does not make any sense. What happened to the polymer concentration in the tank? Where is the analytical data on polymer concentration with time? 

“Root Cause – A fundamental, underlying, system-related reason why an incident occurred that identifies a correctable failure(s) in management systems.” 

“By this definition, a root cause is the most fundamental level in the cause determination, and there is no more fundamental level.” 

“As most root causes are associated with weaknesses, defects, or breakdowns in the management system(s), investigators should look for weak barriers.” 

[Guidelines for investigating process safety incidents, 2019] 

Which of the committee members is trained in root cause analysis? Root cause analysis uses templates. The committee should have attached the templates used for the root cause analysis in support of the claim. Moreover, root cause analysis leads to fundamental level cause determination and not possibilities. This statement is an indicator of professional dishonesty. 

4. “The root cause thus appears to be the lack of experience of LG Polymers India and their Korean principal, LG Chem, in monitoring and maintaining full tanks of styrene that were idled for a long period of several weeks without operation.” (Page 16) 

This is a complete misuse of the term root cause as defined above in 2. No probable ambiguous conclusions form root cause. That is the most fundamental level cause with no further causes beyond it. The report has failed to identify the real root cause as it has not done the root cause analysis as claimed. If they did they should produce worksheets to NGT. 

Styrene is a reactive chemical and its storage requires maintaining safe conditions whether during the plant operations or during shutdowns. As per the report “The unit was permitted for daily maintenance activities during the lockdown period with 15 persons per each shift with a total of 45 personnel working per day.” (Page 14) 

Same regular checks for temperature and polymer content, the recommended standard practices for storage internationally had to be followed during lockdown also as during operational days. It does not call for any additional expertise but only attention.

The report has not recorded any such practices followed at the plant during the regular operational periods and has not presented any operating data for the lockdown period. Without presenting any such evidence in support of the conclusion, it becomes a supposition without basis. It is the lack of experience of the expert team in accident investigation that led to such supposition without collecting and presenting evidence 

Cause is the lack of experience or negligence could have been ascertained easily by following the operational data for the lockdown period and earlier. Accident investigation is a scientific process and making conclusions without evidence does not form an investigation. 

Conclusion: 

The report does not qualify as an accident investigation and failed to identify the underlying causes for the incident and the associated systemic causes like work and safety culture, management and regulatory systems that contributed to it. 

The Chemical Safety Board, USA is an independent agency established to investigate chemical process accidents. Its “investigative staff includes chemical and mechanical engineers, industrial safety experts, and other specialists with experience in the private and public sectors. Many investigators have years of chemical industry experience.” (www.csb.gov)  

The team formed is inadequate to the task. With due respect to their professions and their standings in their professions, Process Safety is a different area of knowledge and nobody can become an expert overnight. 

Signed 

Dr K Babu Rao and other members of Scientists for People 

31 May 2020