Table of Contents

A. Overview

B. Introduction

C. Background

D. Statewide Trends
     1. Utility Outlays and Program Savings
     2. Program Participation
     3. Monitoring & Evaluation
     4. Utility Cost to Acquire Savings
     5. Program Projections

E. Conclusions
     1. What Does the Evidence Suggest?
     2. DSM Potential Remaining and the Mandate for Comprehensiveness

______________

Attachment 1 - Utility Program Expenditures and Annualized Incremental MWh Savings

Attachment 1b - Annualized Incremental MWh Savings

Attachment 2 - Electric Utility Demand Side Management Programs Costs and Savings 1992 - 1997

Attachment 3 - Participants with Installations

Attachment 4 - Evaluation Cost

Attachment 5 - Utility Costs per Lifetime MWh Saved

Attachment 6 - Summary of DSM Programs - Program-to Date Totals - All Programs

Attachment 7 - Vermont Electric Utility DSM Accomplishments, History and Current Trends

HISTORY AND CURRENT TRENDS

A. Overview

As of December 31, 1997, Vermont electric utilities filing 1997 Demand Side Management (DSM) Annual Reports had spent over $75 million for DSM programs since 1991.⁽¹⁾ These programs have reduced Vermont's annual electric use by 249,355 MWh or 4.7%, and they represent a reduction in the state's peak demand of approximately 56 MW of capacity (roughly the size of the McNeil plant in Burlington) or 6%. Stated another way, these programs produced enough electricity savings on an annual basis to serve 34,600 Vermont homes, or about 15% of the state's households.

Since the early 1990's a significant portion of the state's growth in electric energy use has been met by DSM programs. Savings from these programs have been achieved at an average utility cost of 2.4 cents per kWh, less than today's average market price for electricity. This is a significant achievement.

Despite the accomplishments to date, however, if current trends continue, the level of savings attained in Vermont over the first six years of DSM program delivery will not be repeated in the next six years. This is a serious concern since Vermont law requiring DSM programs has not changed and since DPS studies have shown that there remains a considerable amount of cost effective energy efficiency potential that could be a valuable resource for helping to meet Vermont's energy needs in a cost-effective, environmentally sound manner.

The DPS May 23, 1997 plan, The Power to Save: A Plan to Transform Vermont's Energy- Efficiency Markets (The Power to Save), estimates there are cost effective energy savings for 1998, from six core programs alone, of 35,578 MWh at an average cost of acquisition of 2.4 cents per kWh.⁽²⁾ Thus, the potential savings from these core programs alone is about 25% greater than the reported 28,337 MWh saved for all utility DSM programs in 1997. The Power to Save also estimates that there is an additional economically achievable potential DSM resource consisting primarily of retrofit savings not targeted by these six core programs of about 1,315,000 MWh available in 1998.

B. Introduction

The purpose of this report is to identify statewide trends in DSM performance by Vermont utilities through December 31, 1997. The report can serve as a basis for understanding past statewide efforts to increase the efficiency of electric energy use in Vermont and where these efforts may be headed or should be directed.

The report is based on data submitted to the Vermont Public Service Board in the utilities' DSM annual reports. The trend analysis is based on data from the five largest utilities who made a significant effort to implement DSM programs during this period: Burlington Electric Department (BED), Citizens Utilities Corporation (CUC), Central Vermont Public Service Corporation (CVPS), Green Mountain Power Corporation (GMP), and Washington Electric Cooperative (WEC).⁽³⁾ Data from another of Vermont's six largest electric utilities, Vermont Electric Cooperative (VEC), was not included in this analysis of statewide trends because it did not begin its DSM activity until 1996. Numerous smaller utilities were not included in the analysis, as many of them had limited (some of them very limited) DSM activity during this period.

This report does not attempt to analyze in detail the effectiveness of current DSM efforts by Vermont utilities, and it does not offer recommendations about the direction DSM implementation should take. These matters have been addressed extensively by the DPS in Docket No. 5980, in which the Department has proposed the creation of an Efficiency Utility to implement seven statewide core DSM programs. The parties are awaiting a Board Order in that proceeding.

In addition, the DPS has and will continue to analyze in detail the least cost planning efforts and DSM performance of individual utilities as the need arises in various rate cases and other litigation.

C. Background

On April 16, 1990 and June 6, 1990, the Vermont Public Service Board (PSB) issued its Order in Docket No. 5270 detailing the various components of least cost planning. This Order provided specific direction to all Vermont electric and gas utilities to develop programs and take appropriate steps to acquire all cost-effective demand side management (DSM) resources from their customers. This Order also defined the regulatory framework within which utilities were to meet these obligations. The requirement for utilities to conduct and implement "least cost" service to customers was codified by the Vermont legislature in 1991 under 30 V.S.A. 218c. Also in 1991, the PSB approved a 30-year power supply contract with Hydro Quebec only on the condition that all participating joint owners (16 of Vermont's 22 electric utilities) acquire all cost-effective DSM as prescribed in the April 16, 1990 order in Docket No. 5270.

The April 16, 1990 order in Docket No. 5270 established several principles for utilities to follow in designing DSM programs and stated that the Department and Vermont utilities should direct their efforts to ensure that DSM programs reflect those principles. The Principles are:

1. All programs should be cost-effective, and all cost-effective programs should be implemented.
2. Programs should be comprehensive. They should reach all customer classes and use the opportunity with each customer to obtain not just one particular electricity-saving measure but all cost-effective electricity saving measures.
3. Programs should be large scale. They should be designed not just to achieve current savings but also to create in each utility a new DSM capability as sophisticated as its existing supply-side capability.
4. The developing programs should be monitored and evaluated. Careful evaluation is continually needed to improve services and to document the energy savings arising from programs.
5. A significant efforts should be made to prevent lost opportunities for energy saving.⁽⁴⁾

D. Statewide Trends

A number of trends in statewide DSM implementation have been identified through the analysis conducted by the Department. To a large extent, these trends are driven by the activities of the two largest utilities in the state, CVPS and GMP. These two electric utilities serve 70% of Vermont's 317,600 electric customers, deliver 74% of its annual electric energy use, and collect 75% of the amount Vermont consumers pay for electricity. Eighty-four percent (84%) of the statewide DSM expenditures in 1993 and 69% of the statewide DSM expenditures in 1997 are attributable to these two investor-owned utilities.

     1. Utility Outlays and Program Savings

The predominant trend is a steady decline in utility expenditures and corresponding decline in savings acquired by DSM programs since 1993. (See Attachments 1 and 2.) As discussed earlier, statewide annual utility DSM spending peaked at $19.2 million in 1993 as utility DSM programs were "ramped up" and has steadily declined to just under $7 million dollars in 1997. The figure of $19.2 million was 4.2% of 1993 utility revenues; the 1997 expenditure level of $7 million is only 1.4% of 1997 utility revenues. Savings acquired from DSM programs reached its highest level in 1993 at 59,417 annualized MWh and declined steadily to its 1997 level of 28,337 annualized MWh. The savings acquired by the five utilities whose data was analyzed for this report totaled 745 lifetime Gigawatt hours in 1993 and declined to 339 Gigawatt hours in 1997.⁽⁵⁾

     2. Program Participation

There has also been a decrease in the number of utility customers who participate each year in DSM programs. (See Attachment 3.) In 1993 over 37,000 Vermont residential and business customers received DSM services, while in 1996 that number had decreased to a low of 17,400 and climbed only slightly to 19,900 in 1997. The utility DSM reports indicate about half of all electric customers statewide participated in a DSM program over the five or six years of program implementation. This number, however, may be significantly inflated because the same customer is counted as a "separate" customer for each year (and in some cases, for each program) in which that customer has participated.⁽⁶⁾

     3. Monitoring & Evaluation

While there has been a reduction in all expenditure categories reported by the utilities during the period 1992 - 1997, the reduction in monitoring and evaluation (M&E) outlays is particularly marked. (See Attachment 4.) In 1993, the utilities spent $1.2 million on evaluation efforts (roughly 6% of the total utility costs). In 1997, evaluation costs were less than $43,000 (an amount less than 1% of the total utility DSM costs).

This reduction reflects a major change that has occurred in the utilities' approach to DSM program evaluation. Early DSM efforts were subjected to rigorous, and sometimes very expensive, evaluation studies. The best of those studies provided a wealth of information about participants and nonparticipants. Impact evaluations were designed to assess the actual level of resources acquired by the program, and the well-designed studies were structured to identify additional DSM resources and improve program performance. These evaluations were most often conducted by independent contractors to provide an objective view of utility performance.

These evaluation tools have been largely abandoned by utilities in favor of in-house monitoring efforts or, in some cases, no evaluation at all. While in-house monitoring is an important component of DSM program implementation and can be useful for providing timely feedback, it is not, in and of itself, sufficient for sound management of DSM programs. A reduction in the monitoring and evaluation effort of this magnitude signals a significant retreat on the part of utilities from the concept that DSM resources are a viable, quantifiable, and reliable means of meeting a portion of Vermont's current and future electric needs. The lack of commitment to DSM program evaluation at a time when utilities efforts are declining significantly is a troublesome development.

The Department has suggested a number of ways to increase the effectiveness of monitoring and evaluation efforts in order to resume an appropriate level of M&E effort and still reduce duplicative and undesirable M&E expenditures.⁽⁷⁾

     4. Utility Cost to Acquire Savings

Along with the steady decrease in reported expenditures, savings, and the number of participants, there has been a reported decrease in the utility cost to capture DSM savings, measured as the "utility cost per lifetime MWh saved." The utility reports indicate the "utility cost per lifetime MWh saved" has decreased steadily since 1994 from 2.9 cents per kWh to the 1997 cost of 1.8 cents per kWh. (See Attachment 5.) These reported expenditures are lower across all the major cost categories, such as; administration, incentives, and evaluation. Notably, both the 1994 and 1997 "utility cost per lifetime MWh saved" are at or below even today's depressed market prices of power and are well below the cost of power from newly built generation.

Clearly, securing energy efficiency at a lower price (as reflected in an increasingly lower utility cost per lifetime MWh) is desirable. However, the trend being reported by Vermont utilities may be the result of both desirable and undesirable factors.⁽⁸⁾

It is reasonable to expect some reduction in the cost of DSM acquisition as utilities gain experience in operating DSM programs. Start-up costs are typically a large portion of the costs incurred during the first year or two of program operation and the savings from program activities are lower during the ramp-up period. The large initial investment in infrastructure, the need for a shakedown of the program processes and procedures, and the novelty of DSM concepts for both customers and program staff virtually guarantees that costs of DSM acquisition will be greater in the early stages of program implementation. As the program gains momentum and staff gains experience, costs typically come down. There may also be changes to cost and savings reporting methodologies that result in apparent reductions in the cost per MWh saved.⁽⁹⁾

Another factor in the lower cost per unit of savings could be the movement toward market-driven programs with typically lower customer incentives and away from aggressive retrofit program strategies. Retrofit programs typically utilize audits to identify all cost effective efficiency opportunities available at a customer's home or business and offer incentives designed to acquire all the identified DSM resources at that site. Market-driven programs typically offer a rebate at the time of purchase or use other time-of-replacement strategies that utilize builders, designers, distributors and retailers as the primary customer contact for programs.

To a large extent utilities have justified the retreat from audit-driven retrofit programs by claiming that reduced avoided costs make more comprehensive efforts non-cost-effective. However, a closer look at some comprehensive programs and projects does not support this claim. For example, GMP recently operated retrofit programs requiring on-site visits and analysis in the Mad River Valley at a cost per lifetime Mwh that was competitive in comparison to market driven programs.

While the trend of offering lower incentives may be resulting in less expensive programs, it is also possible that these programs are being delivered in ways that "leave savings on the table." For instance, it may be that certain program designs have appeal primarily to a select group of customers who would likely have installed energy efficient devices even without the program, a group often called "free riders." Equally troubling, it is likely the "market driven" programs currently being offered acquire the bulk of their savings from the "easier to get" efficiency resource and leave the more costly, though still cost effective, resources at thecustomer site. This danger was recognized in the Board's Order in Docket No. 5270, and utilities were warned to avoid it.

     5. Program Projections

One final trend is worthy of note. Despite the historic trends of the "utility cost per lifetime MWh saved," a number of utilities have not reduced, and in some cases have actually increased projections of utility cost per lifetime MWh saved since the peak DSM activity in 1993. Statewide, utilities have consistently projected that their programs would be substantially less cost-effective than the actual implementation they claim (See Attachment 5).

Inexplicably, most utilities continue to project costs and savings for DSM programs that assume utility cost per lifetime MWh savings comparable to, or higher than, the actual experience of 1993. For example, the utilities projected program costs of 2.8 cents per lifetime kWh in 1997, in comparison to actual 1995 costs of 1.9 cents and actual 1997 costs of 1.8 cents. It appears that projections are not based on actual performance, which has an impact on both the planning and actual reporting. For instance when the erroneous cost projections are used for planning, programs could be incorrectly determined to be non-cost-effective. On the other hand, inflated cost projections can make implementation in a given year look more successful than it might actually be.

E. Conclusions

     1. What Does the Evidence Suggest?

There was a rapid retreat from DSM acquisition in 1994 by a number of Vermont's largest utilities despite the existence of ample cost effective DSM potential. This is confirmed by the fact that some utilities have achieved DSM savings of substantially more than the statewide average of 4.7% of annual sales. BED, for instance, has saved 10.6 % of its annual energy needs, and WEC has saved 6.9 % of its annual energy needs.⁽¹⁰⁾

As for the utilities at the other end of the spectrum where little or no DSM resources have been acquired, it is hardly appropriate to conclude that the DSM potential has been exhausted.

Finally, there is growing evidence that an application of least cost planning principles known as distributed utility planning (DUP) has significant potential to increase the value of DSM applications in certain site-specific applications where distribution and/or transmission capacity are constrained.⁽¹¹⁾

     2. DSM Potential Remaining and the Mandate for Comprehensiveness

The data suggest there is significant DSM potential available each year that is not being captured by existing utility DSM programs. There are other indicators that this is likely the case. The annual savings per participant is lower in 1997 than in 1993 for both residential and C&I programs. Residential program savings per participant have fallen by one-third, from an average 600 kWh for 1993 to 400 kWh for 1997. Significantly, the Washington Electric Cooperative average residential savings increased by more than one third over this period, from 800 kWh in 1993 to 1,100 kWh in 1997. WEC continues to operate an aggressive residential program providing direct installation of all cost effective lighting and water heating conservation products at the time of the home visit, plus significant customer incentives for space heating and water heating fuel switches from electric heat to an alternative fuel. The higher savings per residential participant for WEC's program, almost triple the statewide average clearly indicates the large untapped DSM potential available from residential customers statewide.

The C & I program average savings per participant has dropped 20% from 21,100 kWh in 1993 to 16,800 kWh in 1997. To date, only 23% of C&I electric customers have participated in a DSM program.⁽¹²⁾ As discussed earlier, the May 23, 1997, The Power to Save estimates there remains considerable untapped DSM potential available from these customers.

As Vermont looks forward to new sources of energy supply in a more competitive electricity market, it should not forget the contribution already made by DSM, and it should not underestimate the potential it holds for meeting future demand in a cost-effective, environmentally sound manner that can help lower customer bills.

1. The amounts discussed here are those provided as "program-to-date" totals in individual utilities' 1997 DSM Annual Reports and include all Vermont electric utilities except the villages of Johnson, Northfield, & Readsboro; Rochester Electric Light & Power Company, and Vermont Marble Power Division.

2. The Power to Save, filed May 23, 1997 in Dockets 5854 and 5980 before the Public Service Board, proposes six statewide core energy efficiency programs: Commercial and Industrial (C&I) Market Opportunities, C&I New Construction, Dairy Farm Program, Residential New construction, Residential Low Income, and Efficient Products Program. The Power to Save also proposes an additional "Emerging Markets Initiative" that would secure primarily retrofit savings not included in this analysis.

3. BED reported significant program implementation costs incurred for a 12 month period that started in mid-1991; however, the rest of the five utilities whose data was analyzed in this report had primarily planning and program design costs prior to 1992. This analysis generally does not include costs incurred prior to 1992, except where program-to-date numbers are discussed.

4. See Vt. DPS Vermont Twenty Year Electric, 1994, 6-7/6-8.

5. Lifetime energy savings are the estimated amount of savings over the life of the DSM measures, while annualized savings represents the energy saved for a single year.

6. For example, about 45% of the number of residential customers GMP reports as participants to date fall under this category.

7. See Vt. DPS The Power to Save: A Plan to Transform Vermont's Energy-Efficiency Markets, 107-116

8. Lower cost per lifetime MWh may reflect better program design and implementation effectiveness. To the extent the trend reflects more effective marketing, innovative delivery systems, broader acceptance of efficient technologies, an ability to attain high levels of savings with smaller incentives, and the move to market transforming programs it represents real gains in Vermont's DSM implementation capability.

On the other hand, the trend could reflect a retreat from acquiring harder-to-get resources that are still cost-effective. For example, a decline in the comprehensiveness of efficiency services to customers and a retreat from potentially significant but slightly higher-cost retrofit efforts might show up as a lowering of cost per MWh, while masking a declining commitment to achieving DSM savings. Of particular concern is the possibility of "cream skimming" in homes or businesses participating in programs. If this is taking place the remaining resources in a building could be too expensive to return and claim at a later date (for example, the remaining savings in a home might not justify the cost of a return visit). Finally, it may also be that the decline in monitoring and evaluation efforts mentioned above means that savings are being over-claimed.

9. For instance, starting in 1994, GMP made a substantial reduction in the payroll overhead rate charged to DSM.

10. It is important to recognize that these two utilities cannot be assumed to have more DSM potential than others because of some unique feature of their service territories, since BED is the state's most urban service territory, and WEC is one of its most rural.

11. DUP is the strategic integration of small generation, energy storage, and demand side management into the electrical system to cost-effectively obtain both location-specific and system-wide benefits. The benefits can include avoiding or deferring investments in transmission and distribution systems, increasing system reliability, obtaining high energy efficiencies through cogeneration technology, and reducing the risk of generation over-capacity.

12. As stated earlier in this document, this is an optimistic amount. The program-to-date participant numbers include an unknown (and perhaps a substantial) number of repeat participants, such as customers who have participated in more than one year, or in more than one program.

ATTACHMENTS

DSM Programs 1998 - 2003 (Assuming 1997 savings per year for next six years)

Savings:	Energy	170,022 MWh	3.0% of Projected 2003 MWh use
	Capacity	36 MW	3.6% of Peak Capacity
Utility Expenditures	$ 42 million