OPINION OF THE COURT
Yesawich, Jr., J.
These consolidated CPLR article 78 proceedings arise out of a prudence review undertaken by the Public Service Commission (hereinafter PSC) to determine whether costs incurred by Long Island Lighting Company (hereinafter LILCO) in the creation of the Shoreham Nuclear Generating Facility (hereinafter Shoreham) exceed the reasonable and prudent cost of the facility because of LILCO’s imprudence, mismanagement and gross inefficiency. The PSC majority of four Commissioners concluded that some $1.395 billion of Shoreham expenditures were imprudently incurred, and excluded this amount from LILCO’s rate base. Three dissenting Commissioners would have precluded LILCO from recovering at least $1.875 billion in Shoreham costs from its ratepayers. LILCO protests the disallowance and also the PSC’s refusal to reopen the record for the purpose of receiving additional evidence; the Attorney-[138]*138General, the Consumer Protection Board (hereinafter the CPB) and the County of Suffolk urge adoption of the dissenters’ opinion.
The Shoreham project began in 1965 when LILCO’s board of directors approved plans to build a 540-megawatt nuclear power plant on the north shore of Long Island at a cost then estimated to be $124 million; commercial operation was expected to begin in 1973. To date, Shoreham’s cost exceeds $5 billion and it has yet to operate commercially.
In 1979, the PSC became sufficiently concerned with continuing construction schedule delays and the alarming escalation in costs of the controversial nuclear generating facility to initiate a prudence investigation. An extensive examination of LILCO’s management of the Shoreham project, made by the staff of the Department of Public Service (hereinafter Staff), with the aid of outside consultants, led to a hearing which yielded over 11,600 pages of testimony and nearly 2,800 exhibits, and culminated in a comprehensive opinion and order by the PSC issued December 16, 1985, which in large part adopted the recommended decision of the Administrative Law Judges (hereinafter ALJs) who conducted the hearing.
The PSC found that LILCO had committed a series of serious management errors and had failed to deal effectively with pronounced engineering, labor, procurement and regulatory problems (Matter of Long Is. Light. Co.—Phase II, [1985] 25 PSC 5904 [Opn No. 85-23]). In its decision, the PSC cited LILCO’s failure to engage in systematic project planning; confusion of responsibilities, primarily between LILCO and its architecture/engineering firm, Stone and Webster; failure to establish, organize and staff systems to enforce construction schedules or implement cost control; an injudicious shutting down of the project in 1971; failure to resize the secondary containment area when the reactor’s size was increased; reliance upon an unworkable procurement process cycle to purchase materials, supplies and equipment; deficient oversight and monitoring of Stone and Webster’s engineering activities and in the planning and scheduling of the construction; lack of control over, and unreasonably numerous, design changes; inefficient and ill-considered resolution of hydrodynamic load problems; poor labor management, resulting in low productivity; mismanagement of regulatory matters, including a harmful lack of cooperation with the Nuclear Regulatory Commission (hereinafter NRC); understaffing of quality assurance [139]*139forces; and failure to adequately supervise the acquisition of indispensable diesel generators.
The record discloses that LILCO’s difficulties began in 1968 when it ignored the warnings of its vice-president of engineering, who recommended deferral of the Shoreham project because of unfavorable regulatory and economic climates. Instead, LILCO’s board of directors authorized construction of an 820-megawatt nuclear reactor. However, when the reactor was redesigned, the size of the secondary containment area was not enlarged. While there was still room enough for the physical apparatus, space was so cramped that workers were in one another’s way, adversely affecting the pace and cost of construction. There was no evidence that LILCO evaluated the adequacy of the secondary containment’s size prior to concrete pour.
Oversight of the Shoreham project was initially delegated to Stone and Webster. LILCO retained Stone and Webster on the recommendation of its executive vice-president, whose son-in-law was the Stone and Webster representative credited with obtaining the Shoreham contract. Stone and Webster received a substantial cost-plus contract without prior approval of the board of directors, a violation of LILCO’s policy on contract commitment.
In August 1971, two weeks after a Federal court decided that the Atomic Energy Commission was subject to the National Environmental Policy Act of 1969 (Calvert Cliffs’ Coordinating Comm, v United States Atomic Energy Commn., 449 F2d 1109), LILCO ordered Stone and Webster to halt its engineering and procurement work indefinitely, despite having a goal of completing 80% of the engineering work without interruption; at that point engineering was only 55% complete. The project was effectively shut down until LILCO finally obtained its construction permit in April 1973. When planning was resumed in September 1972, Stone and Webster assigned new, relatively inexperienced engineers to the project. Because of this, coupled with the engineering hiatus, the project fell far behind schedule; engineering drawings failed to keep ahead of construction, LILCO lost priority with its suppliers leading to procurement problems, and time that could have been devoted to bringing designs into compliance with new regulations was dissipated. Again, LILCO did not perform the studies needed to appraise the advisability of a shutdown, nor did it ascertain whether engineering was sufficiently advanced to support construction once the project resumed.
[140]*140Construction began in June 1973, but incomplete and inadequately checked engineering drawings hampered progress. In an effort to reduce its grave state of delay, LILCO developed an expedited construction schedule which contemplated two work shifts, but construction unions refused to work the second shift called for in the plan. Not only did this schedule fail almost immediately, but worker morale was badly deflated by goals now impossible to achieve. Nevertheless, LILCO retained the unworkable schedule, causing the project to slip even farther behind.
With ample basis for doing so in the record, the PSC found that the project’s rate of progress continued to worsen due to LILCO’s failure to enforce monitoring and control techniques, required by contract of Stone and Webster, which would have enabled LILCO to pinpoint problem areas and identify solutions. Instead, LILCO knowingly left such matters as low labor productivity, engineering and procurement delays, quality assurance and reactor building congestion largely uncorrected.
In May 1976, to overcome the construction gridlock, LILCO established the Unified Construction Organization (hereinafter UNICO) consisting of its own and Stone and Webster’s management personnel. UNICO, a construction monitoring group, was to provide coordinated leadership. When little improvement was had, a task force of select LILCO and Stone and Webster personnel was formed. The task force, after assessing the situation, reported: "There are no generic constraints which caused the poor production and productivity which are being experienced in the piping area. Neither is the cause individual incompetence. Rather, the causes are comprised of a combination of poor planning, inadequate direct supervision, cumbersome procedures and organizational arrangements, inaccurate controlling and reporting systems, a lack of organizational leadership, and a serious antagonism between Stone & Webster and Courier [the piping contractor].” The task force assumed responsibility for a portion of the reactor building and through careful scheduling and close supervision obtained dramatic improvements in piping production, then the chief construction activity. In December 1976, the task force was disbanded and its report was referred to Stone and Webster, which inexplicably did not implement similar practices. Within a month thereafter construction progress fell below pretask force levels.
In 1977, in response to the suggestion of an outside manage[141]*141ment consulting firm that neither LILCO nor Stone and Webster was providing the leadership that Shoreham required, LILCO undertook to superintend the construction. Unfortunately, LILCO’s supervisory personnel did not possess relevant nuclear construction experience and they continued to rely primarily on Stone and Webster’s inadequate reporting and management systems. LILCO also experienced chronic labor productivity problems. In 1977 a work force sampling revealed a direct work percentage of 21% (compared to an industry average of 31.5%) with 8% of hours wasted through late starts and early quits and 42% lost to waiting at the job site. LILCO’s proposed solution to this dilemma was to emphasize production over productivity. A "human wave” of additional workers was hired to enhance production, but this hope did not materialize because of the already congested and poorly coordinated worksite; indeed, the piping contractor, alluding to the lack of sufficient facilities and engineering support, announced its intention to reduce its work force by 200 employees. By October 1978 LILCO’s estimate of the costs of the project, which was periodically being revised, was now $1.337 billion.
Although start-up testing of individual systems of the reactor began in 1976, a dispute between construction and start-up organizations, punch list problems and belated engineering efforts to bring about design changes frustrated any meaningful testing until 1981. Of the items to be tested were three emergency diesel generators (hereinafter EDGs) to be used to supply on-site power to the reactor shutdown equipment in the event of an emergency (see, 10 CFR part 50, Appendix A, criterion 17). The EDGs, received in June 1976, were tested in August 1983. Upon preoperational testing, the crankshaft of one EDG severed and inspection of the other two units revealed cracks in the crankshafts, connecting rod bearings and 23 of 24 pistons. The crankshafts failed because they were undersized for the power levels at which they were to operate; cost of repairing and replacing the EDGs was estimated at $95 million. Rate recovery of these costs was disallowed by the PSC due to LILCO’s imprudence in selecting a diesel manufacturer which had never previously qualified a diesel engine for use in a nuclear power plant, and yet neglecting to take adequate precautions to insure proper monitoring of the EDGs’ preliminary tests. The failure of the EDGs prevented low power testing, a prerequisite to an operating license, until June 14, 1985, when the NRC approved the repaired EDGs as [142]*142Shoreham’s backup power source. By the end of 1984, Shore-ham’s cost, as predicted by LILCO, was expected to be $3.87 billion.
For lack of an NRC-approved radiological emergency response plan (hereinafter RERP), LILCO has yet to obtain an operating license. LILCO charges this to the vigorous opposition of Suffolk County (once a contractual partner in the preparation of Shoreham’s RERP) and its refusal, and that of the State, to participate in LILCO’s proposed plan.
As a result of the prudence review, the PSC determined that LILCO’s imprudence was widespread and encompassed planning, supervision, construction, cost control, engineering, responses to regulatory quality review and the faulty backup generators. The PSC was persuaded, by Staff, that the pervasive nature of the mismanagement, the absence of reliable records on the Shoreham project (a symptom of the poor management), and LILCO’s inability to produce an issue-by-issue quantification of the consequences of its imprudent behavior justified Staff’s reliance upon comparative analyses to fix the extent to which costs of the Shoreham plant should be excluded from LILCO’s rate base. Using "industry average” comparisons, two figures were extrapolated: excessive engineering hours (2.9 million) and excessive construction hours (7.6 million), from which appropriate cost disallowances were then computed. Imprudent construction delay was also estimated using comparisons. The cost of repairing and replacing the EDGs and the costs resulting from delay were calculated directly (i.e., based on actual costs rather than comparisons). Finally, the various delay costs were halved to account for the fact that licensing of the Shoreham plant for commercial operation was equally impeded by factors beyond LILCO’s control, primarily the opposition by Suffolk County and the State, and unusually earnest scrutiny by regulatory bodies such as the NRC. The end result of the prudence review was the PSC’s disallowance of the following items:
Excessive Engineering Costs Excessive Construction Labor Costs
EDG Repair and Replacement One half of All Delay Costs Total Disallowed Costs
$ 140.8 million
$ 399.8 million $ 95.0 million1
$ 759.0 million $1,394.6 million.
[143]*143Six months after the PSC’s opinion and order were handed down, and approximately IV2 years after the record was closed, LILCO moved to reopen the proceeding to offer additional evidence primarily to quantify the cost impact of the imprudent acts identified by the PSC. In denying that motion, the PSC observed that LILCO had ample opportunity to present this evidence during the hearing.
LILCO instituted two CPLR article 78 proceedings, one directed at annulling the prudence order and the other challenging the denial of its motion to reopen the record. The Attorney-General brought a CPLR article 78 proceeding, contending that the full amount, and not merely 50% of the delay costs stemming from LILCO’s imprudent management should have been disallowed from LILCO’s rate base. A CPLR article 78 proceeding instituted by Suffolk County and the CPB seek similar relief, but maintain in addition that inquiry into Shoreham delay costs after March 1, 1986 (the date as of which the PSC found the failed EDGs no longer chargeable with responsibility for delay costs), should not have been foreclosed. These four proceedings were consolidated and then transferred to this court.
THE LEGAL STANDARD FOR DETERMINING IMPRUDENCE
LILCO’s most fundamental criticism is that the PSC, after enunciating the proper standard for determining prudence, abandoned that standard in favor of one which amounted to strict liability applied by hindsight, and thus placed an unduly heavy burden on LILCO. We are unpersuaded.
The PSC’s power to investigate the propriety of costs incurred by a utility derives from its duty to set just and reasonable utility rates (see, Public Service Law § 66 [12]; § 72). That duty translates into the task of establishing which utility costs should be shouldered by the utility’s shareholders rather than its ratepaying customers (Rochester Gas & Elec. Corp. v Public Serv. Commn., 51 NY2d 823, 825, appeal dismissed 450 US 961). It would be neither just nor reasonable for a utility’s customers to bear the cost of inefficient management or poor planning (Matter of St. Lawrence Gas Co. v Public Serv. Commn., 42 NY2d 461, 465).
Prudence, an essential constituent of utility regulation, is determined by judging whether the utility acted reasonably, under the circumstances at the time, "considering that the [144]*144company had to solve its problems prospectively rather than in reliance on hindsight” (Matter of Consolidated Edison Co., PSC Opn No. 79-1, at 6 [Jan. 16, 1979]). Ultimately, however, ”[t]he burden of proof is upon the utility whose rates, rules and regulations relating thereto, charged or proposed to be charged, are being considered”, to justify its conduct (16 NYCRR 61.1; see also, Public Service Law § 66 [12]; § 72).
Historically, utility expenditures initially have been assumed to be exercises of reasonable managerial judgment (see, Matter of New York City Hous. Auth. v Public Serv. Commn., 23 AD2d 277, 280, mod 17 NY2d 246, citing Missouri ex rel. Southwestern Tel. Co. v Public Serv. Commn., 262 US 276, 289 [Brandeis, J., concurring]). Given this starting point, Staff is obliged to demonstrate a tenable basis for raising the specter of imprudence before the utility can be called upon to defend its conduct; this was the approach taken here. Once Staff has come forward with evidence of imprudently incurred costs, the burden shifts to the utility to prove either that its actions were reasonable or that they accounted for no additional costs.2 Though the PSC did not, with every finding of imprudence, suggest a specific alternative that LILCO should have adopted, the record contains ample evidence substantiating inordinate expenditures due to imprudent management.3 The burden imposed on LILCO in no way exceeded that prescribed by Public Service Law § 66 (12) and § 72.
While there is language in the PSC’s opinion lending force to LILCO’s criticism that the PSC may have engaged in some hindsight analysis in arriving at its finding of imprudence, such as giving consideration to whether a cost could, rather [145]*145than should, have been avoided, on the whole, the opinion manifests application of the proper standard of review. It is readily apparent that the PSC familiarized itself with the circumstances which existed at the time of LILCO’s alleged imprudence and from that perspective isolated each major phase of the Shoreham project and then painstakenly examined the decisions made by LILCO officials. The PSC thus reached its conclusions concerning the reasonableness of LIL-CO’s decisions, not retrospectively, but as it was required to do, from the vantage point of the circumstances confronting LILCO at the time. That analysis revealed that LILCO recognized, but ignored or unreasonably mishandled, existing and future problems with the Shoreham project.
QUANTIFICATION OF THE COSTS OF IMPRUDENT CONDUCT
In its original order establishing the prudence investigation, the PSC provided LILCO with a list of specific questions and requested LILCO to respond with quantitative evidence. When LILCO was unable to produce reliable data, Staff developed an innovative approach for quantifying LILCO’s alleged imprudence. This consisted of focusing on the cumulative effects of LILCO’s putative mismanagement in key cost categories (engineering, construction labor, schedule delays and diesel generators) and comparing Shoreham’s statistics in these areas with corresponding statistics from other nuclear power plant projects. LILCO strenuously objects to both the accuracy of the comparisons and the causal relationship between its acts and the cost figures obtained by this methodology.
Since the comparison method plainly lacks precision, it hardly inspires confidence in its accuracy. LILCO convincingly demonstrates many respects in which the comparisons are not really apposite, weighing against LILCO, as they do, unfairly. Responding, the PSC adverts to aspects of the comparisons which too generously extenuate LILCO’s shortcomings, making the final assessment of wasted dollars conservative. LILCO further protests that no causal connection is made between the costs disallowed and particular findings of imprudence as there would be had the PSC assessed the cost of each instance of imprudence. In this regard, LILCO notes that despite PSC rulings which reversed findings of imprudence by the AUs, no parallel reduction of disallowed costs was made.
Nevertheless, the fact remains that LILCO was the only party that might have been able to provide the information [146]*146necessary for an issue-by-issue estimation of the costs of LILCO’s improvidence. LILCO was not forthcoming with that information, if indeed it was able to produce it at all. The cost impact of imprudence, though exceedingly difficult to calculate, if only because of the daunting complexity of the very project itself, was made even more arduous to measure by reason of LILCO’s failure to prudently monitor how, when and at what cost work was done. Staff explained, and the PSC accepted Staff’s rationale, that a number of factors made issue-by-issue quantification of the cost of LILCO’s conduct, found to have been imprudent, virtually impossible. For example, management’s derelictions were so ubiquitous and, in many cases, ongoing that they overlapped; in other situations the effect of one error was multiplied by still others. This massive intertwining of cause and effect made issue-by-issue computations infeasible.
The PSC has wide latitude in deciding how to determine just and reasonable rates, especially in matters requiring technical expertise (Matter of Abrams v Public Serv. Commn., 67 NY2d 205, 211-212). Admittedly, industry average comparisons are clumsy vehicles for gauging costs, but given that multiple comparisons4 were employed and that LILCO, the party charged with the burden of showing that the PSC’s approach to quantification resulted in rates that are neither just nor reasonable (see, Matter of St. Lawrence Gas Co. v Public Serv. Commn., 54 AD2d 815, affd 42 NY2d 461), offered no alternative analysis, it cannot be said that the PSC’s determination lacked a rational basis or reasonable support in the record (see, Matter of New York State Council of Retail Merchants v Public Serv. Commn., 45 NY2d 661, 672).
A related issue is LILCO’s desire to reopen the record to enable it to offer its own quantification evidence. The decision to refuse to reopen the record is within the PSC’s discretion [147]*147(New York Tel. Co. v Public Serv. Commn., 36 AD2d 261, 268, mod 29 NY2d 164). LILCO had been asked to furnish quantitative analysis of the effect of numerous factors on Shore-ham’s costs in the order instituting the prudence investigation in May 1979. During the ensuing five-year investigation, LILCO substantially failed to comply with that request. Granting LILCO’s motion long after the record was closed, to enable it to introduce its own quantification evidence, would only encourage utilities to stonewall the PSC. Inasmuch as LILCO had ample opportunity to present this evidence during the hearings, the PSC’s refusal to reopen the record was not an abuse of discretion.
RESPONSIBILITY FOR FAILURE OF THE EMERGENCY DIESEL GENERATORS
LILCO was held responsible for the EDGs’ failure.5 The record evidence is that the EDGs failed because their crankshafts were not designed for the use to which they were to be put. Stone and Webster was aware that this was a potential problem, yet neglected to adequately oversee tests that it had requested to resolve its concerns.
In an attempt to insulate itself from accountability for the EDGs’ failure, LILCO points to its delegation of oversight for procurement and manufacture to Stone and Webster and the concealment of design defects by the EDGs’ manufacturer, Transamerica Delaval (hereinafter Delaval). We share the PSC’s view that LILCO is indeed answerable and that the simple expedient of a utility delegating its responsibility to an agent cannot be used to shield it from accountability. Implicit in the admonition that a public utility is bound to maintain just and reasonable rates is that it must do so through prudent conduct. It is the utility that is beholden to its ratepayers for a judicious investment of its resources, and that obligation cannot be shirked by entrusting it to agents or independent contractors and then blindly relying on them (see, Matter of Central Hudson Gas & Elec. Corp., PSC Opn No. 82-9, at 5 [May 11, 1982]). Consequently, Delaval’s purported deception is simply not relevant in these proceedings, for the PSC’s determination is not directed at allocating liability between Delaval and LILCO, but at separating costs to be [148]*148borne by LILCO’s shareholders and ratepayers, respectively (see, Rochester Gas & Elec. Corp. v Public Serv. Commn., 51 NY2d 823, supra; cf., Long Is. Light. Co. v Transamerica Delaval, 646 F Supp 1442, 1452).
Furthermore, there is creditable evidence in the record that LILCO itself acted incautiously with regard to its acquisition of the EDGs. It awarded the contract for the EDGs to the only biddpr with no previous nuclear experience and no nuclear quality assurance program, and whose generator was an untested prototype. In doing so, LILCO rejected the modestly higher (4.8%) bid of a manufacturer that had already qualified its machines for use in nuclear power stations. Having selected the least experienced manufacturer, LILCO then did nothing to ensure compliance with the contract and when a fundamental test of the crankshaft, beyond that called for in the contract, was required, LILCO remained oblivious to the purpose for and results of the test.
APPORTIONING DELAY COSTS
The costliest disallowance ordered by the PSC is that attributed to delay costs; these were arrived at by dividing the project into two phases. The first delay phase considered by the PSC covers the period from August 1, 1982, when the record indicates that Shoreham should have entered into commercial operation, and April 1, 1984, the date LILCO projected that it would achieve commercial operation. LILCO’s amenability for this phase of the delay stems from engineering and construction inefficiency (although only construction delay was calculated to avoid double counting) and the failure of the EDGs. The PSC also found that even if these problems had not surfaced, operation of Shoreham would still have been hindered and necessarily delayed by factors beyond LILCO’s control, primarily, unusually severe regulatory scrutiny, changing regulatory standards, and the militant hostility of Suffolk County, the State and those sharing kindred interests. Accordingly, although Staff recommended disallowing $610 million for costs incurred during this 21-month period, the PSC assigned only one half, $305 million, to LILCO’s maladministration.
The second delay phase addresses the consequence of LIL-CO’s diesel generator imprudence, exclusive of the $95 million disallowance occasioned by the need to repair and replace the EDGs. This phase encompassed the period between April 1984 [149]*149to March 1, 1986, after which delay costs were viewed by the PSC as being the product solely of lack of an approved emergency plan. LILCO’s share of the delay costs imputable to the inadequate EDGs was calculated at $524 million.6 This amount also was one half of the actual delay costs. Again, the PSC found the opposition to an RERP, "enhanced regulatory scrutiny, and the extraordinary intervention affecting the Shoreham project” were such that halving this delay adjustment between LILCO’s ratepayers and shareholders was appropriate.
Unquestionably, LILCO’s mismanagement and the antagonism to Shoreham’s emergency planning are interrelated and concurrently generated these delays. But apart from the AUs’ bald and conclusory declaration that "about half of these schedule delay costs were due to the special impact of Shoreham’s unique licensing history and not to factors within the control of LILCO”, there is no further elaboration in their decision or in the PSC’s decision as to why these costs were, or should be, shared evenly. The PSC’s very considerable discretion in choosing how to determine rates does not extend so far that it can turn to speculation or forgo adequately setting forth the basis in the record for its conclusions (see, 300 Gramatan Ave. Assocs. v State Div. of Human Rights, 45 NY2d 176, 180; Matter of Montauk Improvement v Proccacino, 41 NY2d 913, 914).
The relationship between LILCO’s mismanagement and the resistance by Suffolk County and others, as causes of the delay, is obvious and demands further explanatory comment before it can be said that these delay costs have been fairly apportioned on a rational basis. At certain times one or the other may have alone been responsible for portions of the delay, and at other times the two causes coincided. Each seems to have had a ratchet effect on the other. Mismanagement delays gave opponents time to intervene and muster opposition, shook public confidence in the project and stimulated heightened administrative scrutiny, particularly that of the NRC and the PSC. On the other hand, opposition-induced delays exacerbated the effects of LILCO’s diffuse management practices, and Suffolk County’s turnabout on emergency plan[150]*150ning appears to have thwarted successful completion of a RERP. We realize that quantification of these catenating causes will not be easy, but something more than a visceral reaction to the evidence is required.
In sum, after considering the parties’ many arguments, we find there is substantial evidence in the record to warrant confirming the PSC’s decision to disallow excessive engineering costs of $140.8 million, excessive construction labor costs of $399.8 million, and the EDG repair and replacement costs tentatively set at $95 million. As for the PSC’s insufficiently explained decision to allocate 50% of the delay costs to LILCO, we withhold decision and remit this aspect of the matter to the PSC for a more elaborative determination, including the delineation of specific findings which underlie the PSC’s conclusion.
Mahoney, P. J., Main, Mikoll and Harvey, JJ., concur.
Determination modified, without costs, by withholding decision as to that portion which allocated 50% of the delay costs to Long Island Lighting Company; matter remitted to the Public Service Commission for further proceedings not inconsistent herewith; and, as so modified, confirmed.