Comment Summary on LNG Annual and Incident Forms

OMB Control # 2137-0522

One Rule Comments Specific to LNG Annual and Incident Forms

Comments on LNG Annual Report Form

Q1. General Comments

AGA, NiSource, INGAA, and Southern LNG (SLNG) commented that much of the data that would be reported on this form duplicates data currently submitted semi-annually to FERC, to the US Coast Guard (USCG), or to PHMSA as a result of incidents. MidAmerican noted that terminology is inconsistent between this form and the LNG incident report form. MidAmerican also cautioned that “incidents” should not be referred to as “events.” BG&E contended that this information is unnecessary given that LNG facilities are static and do not expand or change over time as do pipelines.

Part B – System Description

MidAmerican questioned the relationship of information across a given row of this part. They noted that plants can be installed on different dates, in different states, and can have significantly different storage capacities. MidAmerican also noted that this part of the proposed form included an apparent formatting error in that lines denoting rows in the table do not extend across all columns.

Part C – Releases in Past Year From Incidents and Safety-Related Conditions

BG&E contended that PHMSA should not collect this information on annual reports because some of it relates to economic issues (e.g., insulation performance), rather than to safety issues. BG&E recommended that information related to incidents should be collected via the incident report form rather than annually.

MidAmerican suggested we reformat this part because it is difficult to follow for operators trying to categorize releases by cause.

Part D – Other Events

AGA, NEGas and NWN recommended deleting this part. These commenters noted that other events are, by definition, not incidents. At most they are “near miss” events of limited relationship to safety and about which it will be difficult to collect consistent data. MidAmerican, NWN, and DOMAC cautioned that events reported on incident reports should not be reported again on this form, contending that summaries prepared for a different form at a different time are almost certain to result in confusion and apparent inconsistencies. MidAmerican, SWGas, and Paiute noted that this part is vague and needs clarification; they commented that several of the listed events appear to be subsets of emergency shutdown. NiSource and DOMAC recommended deleting rollovers and security breaches because these are not safety-significant events. MidAmerican maintained that both terms require better definition, noting that LNG is in constant rollover in tanks due to thermal gradients and suggesting that false activations of security systems/detectors should not be included as security breaches.

Instructions

TPA noted that the instructions need to address electronic filing and the requirements to apply for alternate reporting methods.

A1. PHMSA Response to General Comments

Many LNG facilities under PHMSA jurisdiction do not fall under the jurisdiction of either FERC or USCG and do not report to those agencies. PHMSA thus cannot rely on data reported to those agencies for a complete understanding of the LNG facilities for which it is responsible. PHMSA understands that LNG facilities experience less year-to-year change than do pipeline facilities and that it would be an unnecessary burden for LNG facility operators to report the same data on consecutive year’s forms. PHMSA has revised the LNG annual report form so that operators may report there has been no change from the data reported in the prior year. In that event, operators need not complete the remainder of the form.

PHMSA agrees that there was a formatting error in Part B of the form that was posted in the docket for comment. Lines denoting rows within this part should have extended across all columns, but did not. PHMSA has revised the format of Part B to improve clarity. PHMSA considers that this change also resolves the apparent confusion about reporting of dates, locations, capacities, etc., as these now clearly relate to individual facilities.

PHMSA has also revised the final form to change the formatting of Parts C and D. As proposed, these parts were in parallel columns, which appear to have caused confusion. In the revised form, these parts each extend across the entire form, which improves clarity. PHMSA does not agree that events to be reported in Part C (e.g., insulation performance) are solely economic issues with no safety significance. Events to be reported in Part C are releases of gas or LNG that result from these causes. Releases may have safety significance and are appropriately of interest to PHMSA.

PHMSA agrees that events that have been reported as incidents should not be reported again on the annual report, and has revised Part D to eliminate categories that duplicate reportable incidents. PHMSA does not agree, however, that Part D should be deleted because none of the events is of safety significance. The remaining events do not reach the threshold of reporting as incidents or safety-related conditions, but do represent safety issues. They include, for example, situations that would have been reported as safety-related conditions had they not been corrected before the report of such a condition was required. (The safety significance of the conditions is the same as safety-related conditions. The only difference is time to repair). It is important to trend these safety events. Though individually of less significance, trends in their occurrence could reveal safety problems requiring additional regulatory attention. PHMSA has retained “rollover” as an event to be reported in Part D. PHMSA disagrees that LNG is in constant rollover. PHMSA agrees that blending and mixing routinely occur within LNG tanks, but this does not constitute rollover. Rollover is a term commonly understood within the LNG industry to refer to an event in which significant stratification has occurred within a tank and, as a result, significant quantities of liquefied gas suddenly relocate due to differences in density. Rollovers have resulted in damage to storage facilities and are safety significant events for LNG carriers and their unloading operations at import terminals. PHMSA recognizes that improved designs have significantly reduced the frequency of rollover occurrence, but considers events that do occur to be significant and to require reporting. PHMSA has also retained security breaches as an element to be reported in Part D. PHMSA does not consider it necessary to explicitly exclude false activations of security systems given that element to be reported is an actual breach rather than any activation of a security alarm system.

PHMSA has revised the instructions to reflect the requirements to apply for an alternate (i.e., non-electronic) reporting method.

Comments on the LNG Incident Report Form

Q2. Terminology

AGA, NWN, and NEGas noted that some terms used are not applicable to LNG operations but seem, rather, to be associated with pipelines (e.g., rupture of previously damaged pipe).

A2. Response

PHMSA has revised the form and instructions to more accurately refer to LNG facilities and assure that requested elements are relevant to LNG.

Part B – System Description

Q3. DOMAC recommended that the on-line reporting system automatically populate this information with the operator having an opportunity to override or change as needed, and that information being collected for the OPID Registry should make this practical. BG&E commented that operational information is of limited relevance for incidents and suggested deleting this part.

A3. Response

PHMSA is not deleting this part. PHMSA agrees that information in the OPID Registry and reported on annual reports should allow this part to be automatically populated when operators complete an incident report form electronically. We will configure the on-line system to do so. At the same time, some information may change and not yet have been reported to the Registry or NPMS. For example, the status of a facility may change. A mobile facility‘s location may be different than originally reported. For OPIDs with multiple LNG facilities, the electronic system will be unable to identify the particular facility involved in the incident and will populate data for all facilities. The electronic system will thus afford operators the opportunity to change information that is automatically populated, including deleting information for facilities not involved in the incident. This practice will minimize the burden for completing this information, which could prove useful in understanding and following up on incidents.

Part C – Consequences

Q4. DOMAC suggested revising the form to accommodate the possible situation that no evacuation was necessary and that the area was not unsafe, in which case there would be no elapsed time to make the area safe.

A4. Response

PHMSA has revised the form to replace the question concerning elapsed time until the area was made safe to one asking for a timeline of the incident. This avoids the implication that the situation was “unsafe.” PHMSA has retained reporting for evacuations. We have revised the instructions to require that operators complete this information based on their own knowledge or based on reports by police, fire or other emergency responder. If no evacuation was needed, operators enter zero. If an estimate is not possible, operators are requested to describe why in the narrative portion of the form. Evacuation information is collected in this same manner for pipeline incidents.

Part D – Origin of Gas Leak/Problem

Q5. DOMAC suggested that “gas leak” be replaced with “release,” noting that a release may have been in liquid form. BG&E recommended deleting questions related to distributed control systems (DCS), since such systems are not required, the information is of limited value, and it will be burdensome to collect. DOMAC agreed that information concerning DCS systems would be of limited value, noting that such systems do not detect all hazards (e.g., fire).

TPA commented that the list in question 1 of gases potentially involved is unnecessary given that the form is intended for LNG facilities only.

DOMAC suggested revising the title of question 2 in this part from “leak detection” to “hazard detection.” DOMAC also suggested reorganizing the form to place this part before Part C, since an incident begins with a release it would be logical to begin data collection with the origin of the release rather than its consequences.

A5. Response

PHMSA does not agree that references to DCS should be deleted. PHMSA has revised this part to address “computerized control systems,” encompassing computer-based control systems that may be referred to by terms other than DCS. PHMSA recognizes that computerized control systems are not required to be installed in LNG facilities, but also notes that many facilities use such systems. It is important for PHMSA to understand how useful these systems are in identifying incidents. The information required for computerized control systems is very limited – whether one was in place and whether it initially detected the event – and thus not burdensome to report.

PHMSA has retained the list of gases in question D1. The list simply asks whether the incident originated with natural gas, LNG or “other flammable gas.” Other gases are used in liquefaction processes and could be the origin of events that escalate to incidents. The definition of an incident in § 191.3 refers to events resulting in reportable consequences due to a release of “refrigerant gas,” which may include other flammable gases.

PHMSA has not re-ordered the form to put Part D before Part C. While it is true that most incidents involve a release, the definition also includes emergency shutdowns and events that the operator considers significant even though they do not meet the other specified criteria. These other significant events may not involve a release (e.g., security breach). Part C reports consequences, which is why the event constituted an incident in the first place. PHMSA considers that the order of these sections is appropriate.

Part E – Suspected Causes

Q6. DOMAC commented that this part appears to be taken from a pipeline context and does not fit the LNG environment.

A6. Response

We have revised this part to be more applicable to the LNG environment.

Instructions

Q7. DOMAC noted that the instructions refer to Part 192 vs. Part 193 and will require significant revision. TPA suggested that the instructions for Part D, question 2, refer to "how was the release detected" instead of "where the leak/ problem occurred." TPA also noted that the instructions need to address the requirements for reporting by methods other than electronic reporting.

A7. Response

PHMSA has revised the instructions to be consistent with the form as modified. The instructions include an explanation of how an operator must apply to use alternate reporting methods. PHMSA notes its strong preference for electronic reporting, which will be the required method for all reports addressed in this rule. Allowance is made for alternative methods when operators demonstrate that electronic reporting involves undue burden. PHMSA will review requests for use of alternate methods critically to assure that electronic reporting would be truly burdensome before approving an alternative.

3

2137-0522

One Rule Comment-Responses

Specific to LNG Annual/Incident Report

PHMSA F-7100.3-1/PHMSA F-7100.3