Massachusetts Electric Co. v. Department of Public Utilities
Court: Massachusetts Supreme Judicial Court; September 4, 2014; Massachusetts; State Supreme Court
Three utility companies—Massachusetts Electric Company and Nantucket Electric Company (collectively National Grid), NSTAR Electric Company, and Western Massachusetts Electric Company (WMEC)—appeal orders from the Department of Public Utilities imposing monetary penalties for failing to restore service promptly and safely after outages caused by Tropical Storm Irene and a subsequent snowstorm. The utilities argue that the department erred by not applying a prudence standard, that its findings lacked substantial evidence, and that the penalty calculations were arbitrary. The court affirms in part and reverses in part the department's decisions. It determines that the department correctly applied a reasonableness standard and had substantial evidence to support findings against National Grid and WMEC, along with deficiencies in NSTAR's communication. However, it finds insufficient evidence for NSTAR's failure to respond to certain calls, vacating related penalties and remanding for recalculation. The court also confirms that the department made the necessary findings for most penalties, but reverses and vacates penalties against National Grid for specific performance issues during the last days of restoration following both storms, citing lack of substantial evidence for violations during those periods. The department has historically been responsible for overseeing electric companies' storm response and ensuring timely service restoration.
In December 2008, a winter storm caused widespread power outages for customers of Fitchburg Gas and Electric Light Company, leading to a departmental review that identified inadequate service restoration but lacked authority to impose penalties. This changed on November 12, 2009, when the Massachusetts Legislature enacted the "Act Relative to Public Utility Companies," which mandated the department to create regulations for emergency performance standards and authorized penalties of up to $250,000 per violation per day, with a cap of $20 million for related violations.
Utility companies were required to submit annual emergency response plans (ERPs) for review, which must include management contact information, a communications system for customers, provisions for individuals with medical needs, communication protocols with local officials, employee safety measures, deployment procedures for crews, and identification of necessary emergency supplies. Following the act, regulations were established outlining utilities' responsibilities to prepare for and restore service during emergencies, emphasizing the need for safe and prompt restoration.
The department can investigate utility performance during emergencies, either on its own initiative or upon request, and may impose penalties for violations of the established standards. The excerpt also mentions the impact of Hurricane Irene on August 28, 2011, which caused significant outages for NSTAR and National Grid customers, followed by an October snowstorm that exacerbated the situation.
Widespread outages occurred for NSTAR customers from October 29 to November 3, and for National Grid and WMEC customers until November 6, due to a snowstorm. The department initiated separate investigations into each utility's performance, focusing on NSTAR and National Grid's responses to both Hurricane Irene and the snowstorm, while WMEC's investigation was solely about the snowstorm. The investigations included numerous records requests, written testimonies, sixteen public hearings, and several days of evidentiary hearings. On December 11, 2012, the department issued its decisions, largely rejecting the Attorney General's allegations but finding that all three utilities violated their duty to restore service promptly and safely, as mandated by 220 Code Mass. Regs. 19.03(3). Consequently, penalties were imposed: $18.725 million on National Grid, $4.075 million on NSTAR, and $2 million on WMEC. Each utility appealed to county court under G. L. c. 25, § 5, with cases reserved and reported to the full court for oral argument. The standard of review requires the court to examine for legal errors, with the burden on the appealing party to demonstrate the order's invalidity. The department's expertise is given deference, and its decision is upheld unless it is deemed legally erroneous, lacking substantial evidence, arbitrary, or otherwise unlawful. The utilities argued that the department incorrectly applied a reasonableness standard instead of a prudence standard when evaluating storm performance. The prudence standard assesses whether the utility's actions were reasonable given known circumstances, differing from the reasonableness standard used by the department, which does not allow for mere substitution of the department's judgment for that of utility management.
In Attorney General v. Department of Public Utilities, 390 Mass. 208, 229 (1983), it was determined that the Department lacks the authority to override the reasonable business decisions of utility managers regarding prudence. The reasonableness standard does not afford deference to utility management's judgment. A utility meets the prudence standard by adhering to "fair and prevailing utility practice," yet practices deemed common may still be unreasonable if they fail to restore service safely and promptly after a storm. Compliance with industry standards can indicate a lack of negligence but does not eliminate the potential for negligence findings.
The prudence standard primarily applies to rate setting, allowing utilities to recover costs incurred through efficient management but prohibiting recovery of excessive or bad faith costs. For supply costs to be recoverable, they must be shown to be reasonably and prudently incurred. Several factors indicate that applying the prudence standard to assess a utility’s storm performance is inappropriate, as the Legislature has explicitly refrained from referencing it in General Laws c. 164, § 1J, which governs post-emergency service restoration. This omission suggests a legislative intent to hold utility companies accountable for their performance in restoring power after emergencies, particularly following the deficiencies highlighted after the 2008 winter storm. Legislative debates characterized the storm’s aftermath as a "nightmare," criticizing the utilities' preparedness and responsiveness, and emphasizing the need for stronger regulatory oversight moving forward.
Applying the prudence standard contradicts the Legislature’s intent, which was for the department to establish and enforce its own reasonable storm performance standards rather than defer to utility industry norms. The prudence standard would undermine the department's expertise and experience, favoring utility management's discretion. While the department considered industry practices and historical data in setting these standards, it did not intend to adopt a prudence standard. The distinction between evaluating eligibility for rate recovery and assessing the adequacy of power restoration after emergencies is critical; the former concerns financial burdens on ratepayers or shareholders, while the latter focuses on the utilities' obligations to the public during crises. Given the high stakes involved in emergency responses—such as addressing "wires-down" calls and restoring service to critical facilities—a more stringent standard than mere prudence is justified.
Utilities have criticized the reasonableness standard as vague and unascertainable, yet such a general standard is necessary due to the fact-specific nature of storm performance evaluations. The department's standards are informed by prior evaluations, which establish what constitutes reasonable conduct during emergencies. Historical assessments have emphasized the necessity of pre-storm preparedness, thorough damage assessments, and effective communication. Ultimately, the Legislature aimed for the department to hold utilities accountable for storm performance, a goal that would be undermined if the standards were solely based on industry practices. Therefore, the department acted correctly in applying the reasonableness standard to assess the utilities' performance in restoring power safely and promptly.
The utilities contend that the department's findings lack substantial evidence, which is defined as evidence that a reasonable mind might accept to support a conclusion. The assessment of substantiality includes considering evidence that detracts from its weight. If substantial evidence supports the department’s decision, courts defer to the department's judgment and expertise. However, if the evidence suggests no reasonable probability for the department's conclusion or overwhelmingly supports the contrary, the findings may be set aside.
Specifically regarding National Grid, the department identified substandard performance in six areas during storm responses: acquisition and deployment of resources, damage assessment, response to priority calls, outage management system performance, communication with public officials, and communication with the public.
In terms of acquiring and deploying resources during Hurricane Irene, the department found National Grid inadequately prepared. Despite anticipating a level V emergency, National Grid secured only 233 contractor line crews and 286 tree crews, significantly below the required minimum. A request for additional crews yielded only thirty-seven, which arrived late in the restoration process. For the October snowstorm, National Grid initially classified it as a level III event but ultimately faced a level V situation, mobilizing insufficient contractor line crews during the storm's critical period. While National Grid eventually met or exceeded its ERP's crew requirements later in the restoration, the department noted that the ERP's figures were guidelines requiring justification for deviations.
National Grid failed to justify its inadequate preparation and response during the restoration processes following the Irene and October snowstorm events. The department found that National Grid did not secure additional crews early enough and unreasonably expected mutual aid crews to arrive at the start of restoration efforts. It noted that the crews acquired were not deployed efficiently, with an inappropriate distribution of resources despite a greater need in some areas. National Grid contended that the ratio of deployed crews to outages was a misleading measure, but the department maintained its authority to assess this as a relevant performance metric, concluding that substantial evidence supported findings of unreasonable crew acquisition and deployment.
Furthermore, National Grid did not conduct effective damage assessments post-storms, violating established guidelines. Phase I assessments, required to be completed within twenty-four hours, took at least two days after Irene, and phase II assessments were significantly delayed. Following the October snowstorm, phase I assessments were completed within thirty-six hours, but phase II was not completed at all. The department criticized National Grid for not pre-positioning assessors, reallocating them to other tasks, and utilizing inefficient assessment procedures. National Grid's argument that it only needed to begin assessments within the set time frames was rejected; the department clarified that both phase I and phase II assessments were expected to be completed within the specified time limits.
Substantial evidence supports the department’s conclusion that National Grid violated its obligation to act reasonably in restoring service following emergencies due to delays in conducting damage assessments. Utility companies prioritize downed wire reports, with priority one indicating life-threatening situations, priority two involving interruptions to emergency operations, and priority three indicating nonthreatening hazards. National Grid's average response time to priority one calls was 22.6 hours during Hurricane Irene and increased to 46 hours during a subsequent snowstorm. National Grid disputes these figures, claiming the department's calculations reflect the arrival time of the last crew rather than the first responders. The department rejected this argument, noting that National Grid failed to preserve reliable initial response time data, which compromised the integrity of its records. During Irene, the company lacked dispatch and arrival times for nearly 8,700 calls, and failed to provide ETA records for 86% of 1,148 priority calls. Similarly, during the snowstorm, ETA records were missing for about 88% of priority calls. The findings indicate that National Grid's performance in responding to priority calls was unreasonably slow. Additionally, the department criticized the performance of National Grid’s Outage Management System (OMS), which inadequately processed outage information during both Irene and the snowstorm, contributing to delays in service restoration. The OMS experienced a significant slowdown on August 28, affecting users for 15.5 hours, necessitating a temporary system shutdown to address the issue.
National Grid's Operations Management System (OMS) was inadequately prepared for the call volume during a Level V emergency storm, necessitating manual tracking of orders and calls. Although improvements were made post-Irene, including enhanced processing power and web servers, users still faced significant slowdowns, and the system occasionally failed to update or manage work orders. The company asserted that these slowdowns did not impede restoration efforts; however, it was reasonable for the department to conclude that the need for manual processing did hinder operations.
Communication with public officials was severely lacking, with municipal leaders reporting vague, untimely, and inaccurate information during both the Irene and October snowstorms. This ineffective communication hampered local emergency management efforts. National Grid did not prioritize restoring power to critical facilities as identified in its Emergency Response Plan (ERP) despite requests from local officials. The outage and accident reporting notification protocol system (ORP) faced issues, first being shut down due to overwhelming notifications and later failing to notify officials of 239 qualifying events during the October snowstorm. The number of municipal liaisons increased from forty-one during Irene to ninety-three in October, but the effectiveness of these liaisons was undermined by their lack of knowledge about local governance and electric systems. Additionally, the company’s estimated times of restoration (ETRs) were often inaccurate, as they focused on broad areas rather than specific communities, further eroding public trust.
National Grid acknowledged that during the October snowstorm, its initial Estimated Time of Restoration (ETR) provided to many customers was inaccurate and that updates were not issued until after the ETR had elapsed. National Grid contends that the department's findings on communication failures relied on subjective complaints from municipal officials and the public, arguing that dissatisfaction stemmed from the message about the storm's severity rather than the communication itself. However, there is no evidence that the department improperly prioritized these complaints or that they were consistently unreliable. The department also considered other factors, including National Grid's insufficient municipal liaison provision, the liaisons' unfamiliarity with local conditions, the poor performance of its Outage Restoration Plan (ORP) notification system, and the practice of updating ETRs only after they had been missed. The department concluded there was substantial evidence supporting its findings of National Grid's inadequate communication, contributing to failures in safe and reasonable power restoration.
In the case of NSTAR, the department found that the company failed to restore service promptly and safely in three areas: response to wires-down calls, communication with municipal officials, and communication with the public. While NSTAR's response to priority one calls was deemed reasonable during both the Irene and October snowstorms, the response times for priority two and three calls were often excessively long and did not reflect the urgency required. NSTAR responded to 80% of priority two calls and 75% of priority three calls within twenty-four hours during Irene, and 94% and 91%, respectively, during the October storm. However, some priority two calls took up to 74 hours during Irene and 65 hours during the October storm, suggesting a violation of its Emergency Response Plan (ERP) requirement to respond with the next available resource. The department noted that emergency personnel had to wait for NSTAR’s response, sometimes for over twenty-four hours. Nonetheless, it rejected the notion that a twenty-four-hour limit is the maximum acceptable response time for all priority two and three calls, asserting that not all responses over that time constitute a violation of NSTAR’s obligations to restore power safely and promptly.
NSTAR was found to have timely responded to priority one calls and most priority two and three calls within twenty-four hours. The absence of a commitment in the Emergency Response Plan (ERP) to prioritize these calls, along with the acknowledgment that longer response times may be reasonable during a level V event, led to the conclusion that isolated longer response times did not substantiate the claim that NSTAR failed to respond timely to priority wires-down calls. However, the department determined that NSTAR's ineffective communication with municipal officials regarding priority calls compromised public safety, supported by substantial evidence, including testimony from officials about difficulties in obtaining response time estimates and long wait times for NSTAR personnel to address downed wires. Consequently, the department imposed a $2 million penalty on NSTAR for its response delays; this penalty was based on the number of days it took to respond to priority calls during specific events. Since only the finding related to NSTAR's failure to coordinate with municipalities was factually supported, the penalties were vacated and remanded for reassessment based on this narrower scope. The department also found that NSTAR's communication was often ineffective, resulting in inaccurate information being provided to municipal officials, which negatively impacted restoration efforts. NSTAR argued that the department improperly relied on unexamined complaints, but the department maintained its right to consider such evidence in its proceedings, as it is not strictly bound by judicial rules of evidence in adjudicatory settings.
Records and documents used as evidence in proceedings must be part of the official record, although not all complaints regarding a utility’s performance require cross-examination. The department's regulations state that unsworn statements are not valid evidence for decision-making (220 Code Mass. Regs. 1.10(1) (2008)). Despite references to unsworn statements, the department's decisions were based on other evidence. NSTAR argued that the department disregarded testimony from municipal officials during cross-examination, but the department is not required to mention every witness's testimony in its final decision, and it did acknowledge NSTAR's credibility concerns. The department found substantial evidence showing NSTAR failed to effectively communicate with customers during storm events, including erroneous notifications of service restoration.
Regarding WMEC, the department evaluated its performance across thirteen categories, identifying unreasonably long response times to priority wire-down calls during the October snowstorm. WMEC's average response times were 22 hours for priority one, 35.4 hours for priority two, and 41.8 hours for priority three calls, with specific instances highlighting significant delays in dangerous situations. The department concluded these delays posed public safety concerns and did not meet restoration standards.
The department imposed fines of $18.725 million on National Grid, $4.075 million on NSTAR, and $2 million on WMEC for the identified violations. The utilities contested these penalties, claiming the department failed to discuss relevant penalty factors (220 Code Mass. Regs. 19.05(2) (2010)) or adequately justify the penalties’ amounts and durations. However, courts cannot substitute their discretion for that of an administrative agency regarding penalties it is authorized to impose, even if the penalties seem harsh based on the court's evaluation.
In Vaspourakan, Ltd. v. Alcoholic Beverages Control Comm’n, the court emphasized a limited standard for judicial interference with agency discretion, applicable under extraordinary circumstances. The Legislature grants the department broad authority to impose penalties on utility companies for breaches of storm performance standards, capped at $250,000 per day per violation and a total of $20 million for related infractions. Key considerations for penalties include the appropriateness relative to the company's size, the company's good faith efforts to comply, and the degree of control over the events leading to the violation. Although the department must provide detailed orders for judicial review, it is not required to explicitly mention every penalty factor as long as it is evident that the factors were considered. The department's analysis included evaluating the public health significance of the violations and the extent of each company's infractions, factoring in their customer base, financial expenditure on storm restoration, and employee count. For instance, the penalty against National Grid constituted less than 16% of its storm restoration costs. The department acknowledged the severity of storms affecting operations, including unprecedented wire-down incidents. In balancing penalties, the department accounted for the company's efforts to improve following prior incidents, such as reducing a penalty for communication failures in light of corrective actions taken post-Irene. The department also highlighted management's control over violations, noting that National Grid's damage assessment failures were attributed to managerial decisions, including the decision not to upgrade their operational management system despite known deficiencies.
The department considered the factors outlined in 220 Code Mass. Regs. 19.05(2) when imposing penalties, even though it did not discuss each factor individually. As a result, the penalties' amounts are upheld. The utilities argued that penalties were applied daily without identifying specific violations for each day. However, substantial evidence of ongoing violations is required for penalties to be valid. The department is not required to make detailed daily findings but must provide sufficient evidence that violations persisted during the penalty period. Most penalties against NSTAR, WMEC, and National Grid were supported by evidence. Certain instances, however, showed insufficient evidence to confirm that violations continued throughout the penalty duration, leading to the overturning of penalties in those cases.
For National Grid, the department found sufficient evidence for violations during the Irene event, as National Grid failed to acquire the necessary resources per its Emergency Response Plan (ERP). Conversely, during the last two days of the October snowstorm, National Grid had more crews than required by its ERP, and there was no evidence of unreasonable deployment of those crews. Thus, the department’s findings of violation and the associated $250,000 daily penalties for those two days were not supported by substantial evidence and were vacated. Additionally, the findings regarding National Grid's damage assessment performance were also deemed inadequate for imposing penalties during Irene.
The department deemed National Grid's damage assessment inadequate because its phase I and phase II assessments did not meet the required timelines of twenty-four and forty-eight hours specified by the Emergency Response Plan (ERP) guidelines. National Grid initiated its phase I survey on August 28 and completed it on August 30, while its phase II survey began on August 30 and concluded on August 31. Consequently, the department imposed a fine of $200,000 per day for six days on National Grid for its deficient assessment despite the assessments concluding on August 31. However, the department lacked evidence of ongoing deficiencies beyond that date, leading to the conclusion that the violation did not extend into the last two days of the restoration period, resulting in the vacating of $400,000 in penalties.
The case is remanded for orders reflecting the affirmations and reductions of penalties. National Grid’s penalty is reduced by $900,000 to $17.825 million. The department's penalties against NSTAR, totaling $2.075 million for poor communication practices, are upheld. However, a finding of NSTAR's failure to respond to priority calls lacked substantial evidence, leading to the reversal of $2 million in penalties associated with that finding. The case is remanded for appropriate penalties reflecting the supported findings.
The document also outlines the classification of emergency events by utility companies, emphasizing that the storms in question occurred in 2011, thus amendments made to General Laws c. 164. 85B in 2012 are not applicable. The department is authorized to prevent a utility from recovering service restoration costs if it fails to implement its ERP, although rate recovery is not a concern in this matter. Any penalties imposed by the department are to be credited back to the utility's customers as determined by the department.
An aggrieved party may appeal to the supreme judicial court regarding final decisions made by the department. The department identified several performance deficiencies by Fitchburg Gas and Electric Light Company, including inadequate damage assessment, insufficient repair crews during a storm, and ineffective communication with the public and local officials. These shortcomings were deemed a failure to provide safe and reliable electric service and to restore power efficiently and timely.
In contrast, the storm performance of Massachusetts Electric Company and Nantucket Electric Company (collectively, National Grid) was found reasonable in several areas, such as weather forecasting, communication with life support customers, advance planning, vegetation management, and reporting. National Grid anticipated significant customer outages due to a severe weather event and utilized a mutual aid process for resource sharing among utility companies.
However, National Grid failed to provide priority levels for a significant percentage of calls during two major storms and had an outage management system that could not detect nested outages, leading to inadequate restoration information for customers. The department combined these violations for analysis purposes. National Grid's emergency response plan defined critical facilities needing prioritized restoration, categorized into three tiers based on the importance of the services they provide.
Lastly, NSTAR Electric Company's performance was generally deemed reasonable, or there was insufficient evidence to prove otherwise, across various evaluated areas, including weather forecasting, resource acquisition, damage assessment, outage management, and communication with state officials.
NSTAR's response metrics during significant weather events were analyzed, revealing the company addressed all life-threatening priority one calls within two hours and 90% of such calls, including misclassifications, within 24 hours during Hurricane Irene. Similarly, for the October snowstorm, NSTAR responded to all life-threatening priority one calls within two hours and 100% of priority one calls within 24 hours. In level V events, NSTAR’s Emergency Response Plan (ERP) prioritizes response to lower-priority calls based on public safety considerations.
Western Massachusetts Electric Company (WMEC) contested the department's response time calculations, asserting inaccuracies in measuring wire repair times and misclassification issues. However, the department noted WMEC's failure to maintain adequate data concerning wires-down calls, which hindered an accurate assessment of their response times.
Penalties imposed on National Grid included substantial daily fines for various categories, such as crew deployment and damage assessment, totaling millions for both Irene and the October snowstorm. NSTAR faced penalties primarily for its priority wires-down response, communications with public officials, and customer communications, amounting to significant fines across both events.
WMEC was also fined for its priority wires-down response, with the department noting inefficiencies in resource distribution, though some cited issues did not occur during the final days of restoration. Data evaluated did not conclusively indicate unreasonable performance in crew deployment during that period.
