What is the Problem with Demand Response Baselines?

§  Incentive programs pay consumers to reduce their consumption relative to some level.  In order to gauge their reduction, one must estimate the counterfactual-- how much the consumer would have consumed in absence of the event.  There are several distinct problems associated with this:

o   Measuring the counterfactual of what the consumer would have otherwise consumed requires sophisticated statistical or economic models.  However, electric load is a fundamentally extremely difficult variable to explain and even the best models account for only a fraction of this variability.  Given this, although load reductions are typically communicated as static variables, it may be more accurate to describe them in terms or ranges, or atleast static quantities with error margins.  However, this is rarely the case.

o   Setting up baselines in this manner essentially rewards customers for having a higher baseline.  This creates incentive to game the system, attempting to over consume electricity during the days which will be used to calculate their baseline average.  This is exacerbated if customers are paid ahigher price for each KwH reduced than they actually would pay for that same kwh at the retail rates they pay while establishing their baseline.

o   Heavy support for incentive programs may crowd out dynamic pricing by overpaying customers for reductions that do not actually occur

Source: Jim Bushnell "When it comes to DR, is FERC its own worst enemy?"

Brief History of Demand Response Aggregators AKA Curtailment Service Providers

Aggregators have grown in popularity in recent years because of the profitable opportunity to be the middlemen between customers and utilities. Wholesale market DR programs in almost all ISO/RTO's were comprised of "legacy" incentive based DR programs offered by utilities.  In restructured markets with retail competition, non-utility entities began to spring up.  In ISO-NE, NYISO and PJM, rules had to be developed for these load aggregators.  The aggregators began to rely increasingly on incentive based DR programs because they provide ongoing capacity payments, as opposed to price-based programs that were less certain in terms of revenue.  Capacity payments represent a monthly stream of income for those enrolled.  The amount depends on how much the business can reduce and often there are limits, for example the ability to curtail atleast 200 kW of load within a certain time period.  There may be additional amounts paid out for the actual reductions that happen, calculated using the baseline methodology outlined earlier.  If enrolled in a price-based program, then the customer would determine the market price at which they would be willing to curtail load.  Payments are based on the market price for the amount of load reduced for each hour.  



Arranging with an aggregator is often a more attractive option for end use customers because they do not force them to pay penalties for underperformance.  If an end-use customer fails to perform for the aggregator, some may instead withhold future payments.  Others, such as Enernoc, make the fact that they charge no penalties for non-performance into a big part of their sales pitch to potential customers .  Aggregators often provide infrastructure such as advanced metering and customer support that utilities cannot necessarily match.  Another advantage of allowing aggregators into a market is that they often can leverage previous marketing experience in recruiting new customers .




Smaller C&I customers may not meet size requirements to deal directly with the utility.  Additionally, it reduces administrative costs for the utility by allowing them to deal with a single large entity as opposed to many smaller ones 

How Do California Demand Response Cost Effectiveness Protocols Work?

California seems to be taking the lead in thinking about how to evaluate demand reponse programs at a high level as compared to the other states we address below.  It has its own set of Cost Effectiveness protocols developed by the CPUC.  Before this, each IOU had its own set of criteria by which to judge their programs and their methodology was somewhat opaque, driven in part by proprietary models and data.  This is an attempt to standardize and make transparent the process by which these programs are judged.


Protocols are meant for programs that produce measurable load reductions, not permanent load shifting programs (ie energy efficiency) because these programs are not dispatchable.  It may be applicable to rates such as CPP.  


All IOUs must use the publicly available DR reporting template, a spreadsheet model with various preloaded specifications such as avoided energy and capital costs.  IOUs enter in their own data such as load impacts and energy savings.


The framework for the analysis includes four different cost benefit tests that examine the effects of DR programs for different actors.  Listed are also examples of what these costs and benefits are likely to be.

1) Total Resource Cost Test:  This attempts to examine the net effect on society as a whole, where society is defined as the LSE or IOU plus all of its customers. 

§  Benefits:  LSEs avoided cost of electricity, market revenue, environmental benefits

§  Costs: Increased supply costs

2) Program Administrator Cost Test:  This attempts to examine the net effect on the LSE or IOU alone

§  Benefits:  LSEs avoided cost of electricity

§  Costs: Incentives paid, administrative and capital costs

3) Ratepayer Impact Measure:  Examines the effect on Ratepayers

§  Benefits: Avoided cost of supply electricity, Market revenue, other market benefits

§  Costs: administrative costs, incentives paid out, increase in supply costs (where applicable)

4) Participant Test:  Examines the effect on Ratepayers who are also enrolled in DR programs

§  Benefits: bill reductions, incentives earned, tax credits

§  Costs: bill increases, capital and O&M of DR equipment installed, value of lost service (due to reduction in energy consumption), transaction costs

Implications of FERC Order 745 “Demand Response Compensation in Organized Wholesale Energy Market"

Much literature has focused on FERC order 745. Here are some insights on how this ruling will affect DR customers and aggregators.  


Although there are no firm size restraints mandated by the rule, other constraints make it difficult for, say a small residential consumer to participate directly in this large market. This is because,to participate in a wholesale market and become a recognized Load Serving Entity (LSE), one needs to demonstrate credit worthiness (usually by having a large sum of money in the bank), meet certain technical standards (telemetry, energy management systems, software to bid, etc) and any other requirements specific to that market.These factors make it so that aggregators have a much easier participating than would a normal end user of energy.  However, large industrial or commercial users of electricity with energy managers and sophisticated systems may choose to directly participate themselves outside of aggregator programs.

How Are Demand Response (DR) Programs Administered??

Administration of DR programs

Administration of demand response programs occurs at the utility or independent system operator (iso) level[1]. Utilities run these programs for the benefit of their RTO, or ISO to promote grid reliability and for other reasons.  They are usually (but not always) required to offer these programs by their overseeing authority, the public utility commission.  Aggregators function within these boundaries.  That is, at times utilities set up particular programs that the aggregators can participate in, but they are subject to program constraints like any other participant.  The only difference between joining with the aggregator and the utility is the level of service and support one receives.  In all other respects, they are the same programs and thus, the same rules. A confusing extra feature of this arrangement is that at times aggregators will bundle several distinct program options under one brand one and market it as a different program to the end-use customer[2].  It is worth noting that generally PUCs have no authority over the contract between the aggregator and the end use customer, the source of much of the jurisdictional confusion around aggregators.

Demand Response (DR) Program Classifications

Types of DR Programs

The traditional dichotomy is between incentive-based DR programs and time and price-based programs(Albadi & El-Saadany, 2008).For the price-based programs, the load reduction occurs because prices have reached a pre-specified high level.  Incentive programs require the customer to shed load in response to a system-wide events. in reality, the line between these programs is often blurry.

Among the incentive programs, market-based programs are growing in popularitydue to a recent high-profile FERC ruling [3].  FERC has mandated that wholesale energy market operators pay locational marginal price (LMP)[4] to demand response resources, effectively paying them as if they were analogous to traditional generating units such as natural gas plants.This has been seen as a boon to the demand response aggregators who can now enlist customers in DR programs and bid them load directly into wholesale markets, as opposed to being a middleman between utilities and customers.  It has also encouraged a host of programs where customers can bid their loads in themselves.  There will be more discussion of this ruling very soon.