NTSB Safety Recommendation Letter P-84-013 through P-84-014

Page 1Official PDFNATIONAL TRANSPORTATION SAFETY BOARD WASHINGTON, D.C. ISSUED: June 15, 1984 Mr. G. H. Lawrence President American Gas Association 1515 Wilson Boulevard Arlington, Virginia 22209 SAFETY RECOMMENDAT I O N (5) P-84-13 and -14 -----------------_..___________________I_----- A t 11:15 a.m., c.d.t., on July 12, 1983, natural gas escaping under 60 pounds pressure from a crack in a butt fusion joint in a 2-inch plastic gas main entered an apartment building in Clear Lake, Iowa, exploded, and then burned. Two Interstate Power Company employees were injured, one apartment building was destroyed, and the adjacent apartment building was damaged heavily. Damage was estimated at more than $1 million; none of the residents were injured or killed. 1/ - According to the Plastic Pipe Institute and plastic pipe manufacturers, such as Dupont, Plexco, and Nipak, a butt fusion between two lengths of plastic pipe when properly made, should be as strong or stronger than the plastic pipe itself. A butt weld between two lengths of steel pipe carries the same strength definition, however, here the similarity ends. In the field, accurate, reliable, repeatable, nondestructive testing of steel butt welds can be readily and practically undertaken by X-ray, radioactive isotopes (gamma rays), and magna flux. The American Petroleum Institutes (API) Standards for Welding Pipelines and Related Facilities, API Std 1104, Section 6.0, Standards of Acceptability, and Section 8.0, Radiographic Procedure, set the standards of acceptability for size and type of weld defect, the requirements for producing acceptable radiographs, and the qualifications of the radiographers for the work. Over the years, these API standards have been improved to the point where today field radiographic inspection of butt welds on steel pipes, by qualified, motivated technicians, can be classified as a science. However, the field nondestructive tests for plastic pipe butt fusion cannot be classified as a science. The physical appearance of the fusion is the primary nondestructive field test method. However, the physical appearance test is valid only when all other procedures of squaring the pipe, heating the tool, applying the pressure, and holding the pipe have been rigidly adhered to. During the annual fusion qualification tests given by Interstate to its pipefitters/operators wherein the employees' work was given a physical inspection followed by a destructive test, most of the persons passed both. However, one person's work passed the visual inspection, but failed the destructive test; one person's work failed - 1/ For more detailed information, read Pipeline Accident Report-"Interstate Power Company, Natural Gas Explosion and Fire, Clear Lake, Iowa, July 12, 1983" (NTSB-PAR-84/02). 3762-A#

Page 2-2- both the visual inspection and the destructive test; a!ld one person failed the visua inspection and passed the destructive test. Interstate stated that all both aspects (visual inspection and destructive testing) before beir plastic pipe in actual field operation. The Safety Board is concerned appearance of a butt fusion joint is more of an art than a science and t to one supervisor may not look good to another and, indeed, upon destructi fail. As to the July 12, 1983 accident, the report prepared by th stated that the external appearance of the failed butt fusion joint was good, but fusion weld wherein uniform fusion of the plastic pipe surfaces had not around the pipe. The findings in this accident raise the question abou plastic butt fusion joints which may have been installed and have a good e appearance, but which in fact, may be substandard. On December 30, 1970, the Safety Board issued "Special Study of Effects of D Shutting Down Failed Pipeline Systems and Methods of P Shutdown." 2/ Since then, the Board has issued 33 safety recomrnendatio 16 pipeline accident which involved failure to shut down pipelines in a timely manner. In the Introduction to the Special Study, t h e Safety Board stated: In almost all recent pipeline accidents, the delay in shutting down the failed pipeline system has resulted in an increased magnitude of catastrophe. Had the flow of gas or hazardous liquid been stopped soon after the initial rupture, the effects of many accidents would have been minimized or eliminated. With the ever increasing use of pipelines for natural gas and other hazardous materials and the proximity of these lines to expanding populated areas, it is imperative that systems and methods be developed and put to use which will provide for the rapid shutting down of failed pipeline systems. The Safety Board concluded that: By reducing the time required to shut down a failed pipeline system to minimize the loss of material, the hazardous effects to the public, to persons working near a pipeline, and to property can be minimized or eliminated. Equipment and procedures are currently available which, if utilized, could drastically reduce the shutdown delay ci accidents discussed in this study. The Clear Lake accident is another illustration of the unfortunate co failed pipeline facilities are not shut down in a timely manner. While formal, written plan for shutting down its gas facilities, the plan does not address specifically the rapid shutdown of small sections of pipeline, such as the one involved in this accident. Moreover the plan is given only to supervisory personnel a employees who are the first to be dispatched to a gas leak site As a result of its investigation of a pipeline accident in An March 24, 1972, in which three persons died, one person was injured, destroyed, and a third house was badly damaged, 3/ the Safety Board r the Office of Pipeline Safety (now the Research and Special Programs t h e US. Department of Transportation: I 2/ Special Study of Effects of Delay in Shutting Down Failed PipelineMethods of Providing Rapid Shutdown (NTSB-PSS-71-1). - 3/ Pipeline Accident Report--"Washington Gas Light Company Natura Annandale, Virginia, March 24, 1972" (NTSB-PAR-72-4).#

Page 3-3- Amend 49 CFR 192 to require that each operator maintain a log which shows the receipt and handling of each leak or emergency report received. Information concerning the time that the report was first received, that a crew was first dispatched to the scene, that such a crew arrived, and that the condition was considered safe should be included. In addition, each pipeline operator should be required to analyze his performance in responding to gas leak emergencies and reports. Both the logs and the analysis should be made available to State agencies and the Office of Pipeline Safety. (Safety Recommendation No. P-72-42) (emphasis added). It is important for gas companies to encourage reporting and to maintain an effective line of communication with the public since the public is the best source for reporting most gas odors/gas leaks. An effective leak response system must include the immediate logging of telephone calls about gas odors, the logging of the precise information given (date, time, location, and leak description), and the rapid dissemination of this information to the responding crew. Emergency-type telephone calls should have the highest priority and should be handled completely before attention is given to regular business matters. By so doing, the gas company not only can insure the rapid response to a complaint, but it can check on its own efficiency as to when t h e crew was dispatched, when the crew arrived, what conditions the crew encountered, and when the condition was rendered safe. Many gas companies tape record all incoming emergency-type telephone calls as a matter of record and as a means to assess their dispatching efficiency. Unfortunately, in this accident, there was no record of the first or second telephone calls from the apartment resident who reported the gas odor, no work order was printed giving the particulars of the leak in time for the serviceman to take with him when he responded, and there was no accurate record of the time the serviceman arrived at the leak site. Undisciplined handling of gas odor or gas leak telephone calls, as was apparent in this accident, decreases the speed and effectiveness in the response to those calls. Interstate's leak classification system is adequately described in its Inspection, Operations, and Maintenance Plan but was not disseminated to the personnel who respond to the leak odor complaints. In this case, the gas serviceman who first arrived on the scene stated that he did not know anything about a Grade I leak. To his credit, he realized that he had a situation too large and too complex to handle alone and he requested additional help. However, after he had requested help, he did not follow Interstate's Grade I "action criteria" which states he should have evacuated the premises, eliminated sources of ignition, or vented the area. If he had been trained to take these actions and had performed them after he requested additional help, this accident might have been prevented. It is important that all persons who are required to respond to gas odor/gas leak complaints, not just supervisors, be trained and knowledgeable in leak classification and the proper emergency actions. Since October 14, 1970, the Safety Board has issued 88 pipeline safety recommendations to the pipeline industry and to the Research and Special Programs Administration of the U. S. Department of Transportation regarding the development of written emergency procedures to be implemented during pipeline emergencies. These recommendations have addressed a range of suggestions from the receipt and rapid processing of telephone calls reporting gas leaks, guidelines for the first person responding to the accident site, the ventilation of gas-filled buildings, the evacuation of residents from buildings, to the rapid shutdown of the failed facility. The Safety Board's concern is twofold: first, a gas company must have complete, written, emergency plans; and second, gas company employees who may be called upon to respond to emergencies must be familiar with these emergency plans to implement them consistently, promptly, and effectively.#

Page 4Of equal importance witti written Emergency Plans and Inspection, Operations, and Maintenance Plans is the training of company personnel and others to rapidly implement these plans in the event of an emergency. In addition to the 88 pipeline safet recommendations for written emergency procedures, the Safety Board, has issued 7 pipeline safety recommendations since 1969 concerning training of pipeline compan personnel and others. Board experience shows that company personnel well versed in thei emergency and operations and maintenance plans can reasonably be expecte better during periods of emergencies than those who are not. The serviceman sent to investigate the gas odor complaint a t the accident sit responded promptly once he was notified, checked the area outside, correctly assessed th gas leak as being too big for him to handle, and correctly requested additional help, all i a timely fashion. However, having fulfilled the first task, site examination and situati analysis, the serviceman failed to follow through. Although he entered the building wi the apartment resident, he did not carry his gas detectr (CGI) into the building, an hence, he could not accurately determine the amount of gas present. It is not yet an industry-wide practice to equip servicemen with CGIs; however, it is an accepted practice that when servicemen are equipped with CGIs to use them to check the concentration of gas in a building. Interstate's action criteria for Grade I leaks gives examples of when a leak should be considered a Grade I. Two examples mention the concentration of gas inside a building. Without the use of some type of detector, there is no way to accurately determine the concentration of gas. When the serviceman entered the building and determined that, to some degree, gas was entering the building, he failed to warn the residents, to ventilate the building, to eliminate sources of ignition (such as shutting off the electricity), and to evacuate the building. Had the serviceman ventilated the building or turned off the electricity to the building, or both, ignition rnizht not have occurred before the gas crew shut off the gas. After the accident, the ~ ~ ' v ~ ~ e m a n stated that he had not been trained to do these things although, on a general basis, !le Nas aware of the hazards of natural gas through his years of practical experience. Similarly, when the gas crew arrived and found heavy gas odors in the air, they als failed to warn the residents, to ventilate the buildings (or even to check the buildings E gas buildup), and to evacuate the residents. Later, when the gas company superintende arrived at the site, smelled the gas, and was briefed about the leak and the work done t locate the leak, h e too failed to act to eliminate sources of ignition or to evacuate th residents. Between 8:30 a.m., when the serviceman arrived at the apartment complex an 11:15 a.m., when the explosion and fire occurred, several measures could have been take to have eliminated the potential danger at the accident site; however, none wer Therefore, the Safety Board believes that adequate emergency and inspection, operatio and maintenance plans combined with training subordinate company personnel 'n thes areas could have prevented this accident. Therefore, the National Transportation Safety Board recommends that the Gas Association: Disseminate to its member companies the circumstances of the acci in Clear Lake, Iowa, on July 12, 1983, and urge them to reevaluate thei plastic pipe fusion procedures and to check that their responsibl personnel are explicitly following the procedures. (Class II, Prio Action) (P-84-13)#

Page 5-5- Urge its member companies to reemphasize to their responsible personnel the importance of rapidly shutting down failed gas facilities and the importance of evacuating residents, ventilating buildings, and eliminating sources of ignition. (Class II, Priority Action) (P-84-14) BURNETT, Chairman, GQLDMAN, Vice Chairman, and BURSLEY and GROSE, Members, concurred in these recommendations.#