14. Unresolved Ratemaking Issues

The Comparison Exhibit (Ex. 20) identifies 17 unresolved ratemaking issues. In this section we address the positions of the parties and finally resolve eight of these issues. Four of the 17 issues were resolved elsewhere by adopting a standard five-year average of 2000 - 2004. We also address elsewhere in today's decision the other five issues (notice of new supply facility fees, return on equity, emergency generator reporting, low-income programs, and depreciation in working cash).

There are nine separate customer categories for water sales, and there is no agreement on the test year forecast. The parties agree on the method to estimate sales forecasts for all customer classes. For the residential customer class, they agree to use a monthly econometric model that includes time, rain, temperature and monthly dummy variables. For all other customer classes, the parties agree to use a five-year average of recorded sales per customer. One difference is the inclusion of 2004 data by Apple Valley. We have resolved this issue and will include 2004 data in the test year forecast.

Apple Valley requests that the Commission make a specific finding of fact adopting the use of dummy variables as consistent with the intent of the adoption of the New Committee Method in D.04-06-018, the rate case plan decision.50 It believes that the Commission intended to adopt a common forecasting methodology for sales. Apple Valley's concern is that ORA, or a water company, could opportunistically pick the method that produced a result favoring their interest in any given proceeding and then argue the other method under or over-estimates unit consumption.

ORA's sales forecast witness did not prepare ORA's deficiency notice that required Apple Valley to provide a forecast without dummy variables, but in preparing his testimony he was prepared to accept the applicant's data including dummy variables.51 He accepted it here "[i]n this particular case for this particular year or set of years, [because] the use of dummy variables provided the better correlation ... When you do forecasting, you do various scenarios, various forecasting variables. You try various forecasting variables. And I certainly would try dummies as well as without dummies for comparative purposes."52

The proper forum to interpret D.04-06-018 is in the rulemaking, or by a Petition to modify the decision, where all interested parties have adequate notice. (See Rule 47.) For our specific purposes in this decision, ORA's witness correctly noted that our objective is to find the best estimate, and, in this case, ORA believes that Apple Valley's use of dummy variables provides the statistical "best correlation." We will only find that for this proceeding, with the available data, the use of dummy variables provided a reasonable forecast for Apple Valley's test year 2006.

For water sales and operating revenues, we will use the five-year average 2000 - 2004, and we will use the data set that includes dummy variables for test year 2006. Unless subsequently clarified, Apple Valley must include two data sets in its next rate case application: one with dummy variables, and another without.

The parties agree on the basic method to estimate purchased power. The estimate is developed from the ratio of recorded power consumption in kilowatt-hours (kWh) and actual water production in hundred-cubic-feet (Ccf) multiplied by the estimated water production in Ccf. Apple Valley and ORA disagree on the time period that should be used to develop the factor of power consumption per water production (kWh/Ccf). Apple Valley's 2006 test year estimate is $1,159,600 using data from June 2003 through May 2004. ORA's estimate used a three-year average, 2002 - 2004.

There are two components to resolve: the kWh/Ccf factor, and the sales estimate, already adopted above.

Apple Valley's position is that averaging is not an appropriate methodology to use when circumstances have changed. Apple Valley's rebuttal shows that from 2002 to 2004 the kWhs it takes to pump a Ccf of water had increased. This was attributed to the fact "that the pumping levels of the majority of AVR's Wells are dropping."53 Further, this is caused by localized stress which Apple Valley explained has the effect of lowering the pumping levels of wells.54

ORA argues that the aquifer stress is not constant, and it can change from year to year.55 ORA defends its use of the most recent three-year average as an appropriate indicator of the historical trend.

We have a limited choice of either Apple Valley's weather-affected single season of data, June 2003 - May 2004, or ORA's three-year trend, 2002 - 2004. Using three-years shortens the five-year trend we have identified as the preferred statistical device. We will rely on the most recent three-year trend because Apple Valley has not shown that the single season's significant change is likely to be permanent rather than cyclical, and is not dependent on such as factors as climatic conditions varying from year to year. This three-year trend also includes the recent period that Apple Valley would otherwise use with less dilution of recent data than would occur in a five-year trend.

We have applied the three year-average for the production of Apple Valley's wells and energy consumption in kWh/Ccf to the sales/production forecast to derive the 2006 estimated expense of $1,159,200.

This one issue has been partially resolved elsewhere by adoption of the five-year average method using 2000 - 2004 data. This accounts for $2,631 of the difference. The remaining difference of $30,469 is whether Apple Valley requires an additional temporary employee for the customer service department.

Apple Valley argues that customer growth will be approximately 21%, and ORA accepted that growth rate in the test year. Apple Valley believes it will need additional workforce, in the form of a temporary rather than permanent employee, to meet this increased workload.56 We find Apple Valley persuasive that a 21% increase in the number of customers will likely lead to the need for additional workers. We will adopt its forecast for 2006.

Apple Valley's forecast for 2006 insurance expenses is based on the actual policy for 2004-2005 policy-year and then adjusted based upon the recommendations of Park's insurance broker for a 2006 estimate of $811,300.57 ORA's forecast is $705,400, a difference of $105,900, about 13% less.

ORA argues that Apple Valley relies on the "untested hearsay" of its insurance broker, whereas ORA used historical data escalated by the standard CPI-U rates.58 However, ORA did not cross examine the Apple Valley rebuttal witness on insurance expenses, and did not raise a timely hearsay objection to Apple Valley's asserted advice from its insurance broker. ORA had the opportunity to pursue discovery on the underlying data used by Apple Valley to forecast 2006 insurance expenses, and we will not sustain any objection now.

The question before us is whether Apple Valley's forecast met the initial burden of proof and whether it is more or less credible than ORA's alternative calculation to escalate 2005 costs by the CPI-U. First, we find that a company the size of Apple Valley most probably needs to rely on brokers, or perhaps competitive bidding, for highly complex and competitive insurance options. That is, we do not expect most water utilities to have in-house insurance experts capable of independently forecasting insurance costs. Apple Valley of course would have to produce in discovery the normal business records and correspondence that the broker created, in offering a 2006 forecast. In rebuttal, Apple Valley demonstrated that 2000 - 2004 recorded insurance expense grew more rapidly than would have been forecast relying on ORA's published CPI-U factors.59 Thus, we find that ORA's method of escalating 2005 costs by the CPI-U is unlikely to reasonably forecast 2006 expense.

We will adopt Apple Valley's 2006 forecast. We will direct Apple Valley to discuss in testimony and include in its workpapers for its next general rate case the detailed correspondence, and related requests or instructions to the insurance brokers, to expedite discovery. If Apple Valley selects a different method to forecast insurance costs it shall fully discuss and disclose that method instead.

The issues before us on main office insurance are identical to the other insurance issues in operating expense. We find that Apple Valley met its burden of proof and offered the most persuasive forecast for Test Year 2006.

Apple Valley has a gravity-fed irrigation service that is not connected with the potable water system. There is only one customer on that system and rate schedule - a golf course. Apple Valley has not proposed a cost-based tariff because it believes the customer is a by-pass risk if rates are increased significantly. 60 The requested rate is $0.575/Ccf. Apple Valley asserts that it has not offered a cost study in several recent rate applications.61

ORA states it would prefer to formulate its recommendation for gravity irrigation rates on a cost of service study in order to determine the appropriate rates of the gravity irrigation system. But Apple Valley did not prepare a report despite ORA's request.62 ORA therefore extrapolated utility-wide data to formulate a rate of $0.665/Ccf and a revenue requirement of $226,115.

Without a cost study, we cannot determine with certainty the true cost to serve the gravity irrigation customer, nor any contribution to marginal costs by that customer. We must also consider the credibility of the by-pass threat: whether there are alternative service options and whether they are cost effective if rates were raised to full cost of service.

The Commission has a long-standing practice for energy utilities that allows a special contract rate for customers where there is an immediate threat of bypass, i.e., to leave the utility system. For the energy companies, a series of decisions granted special contract rates subject to certain limitations. In D.92-11-05263 the Commission adopted an expedited review process to consider approval for special deals that would allow the utility an opportunity to retain a customer who may otherwise depart the system and leave the utility with stranded facilities. The following criteria were established:64

1. Bypass should be prevented if it is uneconomic, that is, if the customer's cost to bypass is more than the marginal cost of utility service.

2. Where uneconomic bypass is threatened, it should be possible to offer utility service at a negotiated rate that still contributes to the utility's fixed costs.

3. The utility's marginal cost to serve a customer is the appropriate standard to differentiate economic from uneconomic by pass.

4. The use of LRMCs [long-run marginal cost] as floor rates should ensure that long-term contracts generate a positive contribution to the utility's fixed costs.

Apple Valley failed to meet its burden of proof for gravity-fed rates. It made no evidentiary showing to substantiate a credible threat of bypass. It also failed to substantiate that the existing rate of $0. 575/Ccf is a reasonable anti-bypass rate (assuming a credible threat exists) that would meet any bypass avoidance criteria such as those in D.92-11-052 above.

Apple Valley may subsequently file a separate application if it can demonstrate that a credible uneconomic bypass threat exists and the bypass avoidance rate covers Apple Valley's marginal costs to serve the customer. The Commission has also spoken on the appropriate standard for review.65 Apple Valley must show:

1. Imminent customer bypass (customer's sworn affidavit)

2. Bypass would be uneconomic (criterion 3 above)

3. The avoidance rate is reasonable (criteria 2 and 4 above)

We will adopt ORA's rate of $0.665/Ccf as the best available approximation of the cost of service for gravity-irrigation service.

Parties agree that 90% of Apple Valley's unaccounted for water is from its gravity irrigation system. This system consists of a series of reservoirs and pipes that are used to irrigate a golf course and at times operate a trout farm. In its Opening Brief, ORA summarized its proposal as "not requesting heroic measures to reduce this loss but merely a reasonable review of this situation to identify possible low-cost solutions."66 Apple Valley made a substantial presentation in rebuttal (Ex. 15.) indicating that it operates the system in accordance with its contractual requirements and believes it cannot identify or undertake any cost-effective capital improvements.67

ORA essentially asked for information, which Apple Valley should have, and update on a regular basis, to allow a prudent manager to consider possible improvements to the system. Additionally, water should be carefully husbanded and not wasted. We will direct Apple Valley to include a full and complete update of the information in Ex. 15 in its next general rate case. Apple Valley should be prepared to demonstrate that it has specifically reviewed its operations, considered low-cost and no-cost improvements, and perhaps even contract amendments, that could reduce water losses.

ORA suggests that Apple Valley should replace or repair its low-rated pumps to improve pump efficiency.68 ORA did not brief this issue, although Apple Valley did, in addition to its detailed rebuttal testimony. Apple Valley suggests that its pump efficiencies are typical, and that "it may be more cost effective to repair a high producing well that currently has a fair rating than a low producing well with a low rating." In other words, too simple a replacement requirement might result in uneconomic decisions. We agree, but we also agree with the general premise in ORA's recommendation that Apple Valley should examine its operations on an ongoing basis so that in its next general rate case it can support its overall economic efficiency by proposing those expenditures that provide the greatest long-run benefit to the company and its customers. No further order is required at this time.

50 A dummy variable is a numerical variable used in regression analysis to represent subgroups of the study sample. Using dummy variables enables a single regression equation to represent multiple groups.

51 Transcript, p. 147.

52 Transcript, pp. 149 - 150.

53 Ex. 17, p. 2.

54 Transcript, pp. 241 - 242.

55 Opening Brief, p. 5, and citing to Transcript p. 242.

56 Apple Valley Opening Brief, p. 7.

57 Apple Valley Opening Brief, p. 9.

58 ORA Opening Brief, pp. 10 - 11.

59 Ex. 18, p. 2.

60 Ex. 1, p. 64.

61 Apple Valley Opening Brief, p. 20 ff.

62 Transcript, pp. 251, 257.

63 46 CPUC 2d, 444.

64 46 CPUC 2d, 444. 449. Conclusions of Law 1, 2, 4, and 10.

65 See Conclusion of Law 13, D.92-11-052. The standards above are paraphrased to reflect a "rate" and not necessarily a contract, as was envisioned in that decision.

66 ORA Opening Brief, p. 20.

67 Apple Valley Opening Brief pp. 28-29.

68 Ex. ORA 1, pp. 3-8.

Previous PageTop Of PageNext PageGo To First Page