Earlier this month, I spoke at ACI’s Next-Generation Demand Response conference in San Diego, where attendees mingled to consider future trends of demand response (DR) activity in the electricity sector.
In the mid-afternoon of the first day, in the wake of several presentations (including mine) on the state-of-play in U.S. DR markets, I sat on a panel that was convened on the spot to deliberate a question that hung in the air:
“Is Demand Response Dead?”
Reintroduction to Demand Response
Emerging as a force in the electricity industry in the early 2000’s, DR can be viewed as a logical successor to load management (also known as load control) programs of the 1980’s. These programs involved the following quid pro quo between electric utilities and participating customers:
- The utility remotely modulated energy-consuming devices on the customer’s premises – air conditioning, refrigeration, or lighting – to reduce electricity volumes for brief periods at times of highest demand on the electricity system.
- In exchange, the utility provided the customer a discount or credit on the monthly electricity bill.
For the customer, the inherent value proposition of load management is straightforward: economic savings with minimal negative consequences (e.g., room temperatures slightly different from optimal for short durations).
For the utility, the benefit would seem less obvious. After all, what business wants to sell less of its product?
As a regulated monopoly, an electric utility must ensure that it can provide reliable service to all of its customers at all times. And, because of the dearth of storage on the power grid – note that nature doesn’t like keeping electrons still – the electricity system has generally been sized to meet the moments of highest collective demand, plus additional amounts to account for adverse contingencies.
Accordingly, utilities historically built power plants that ran for very few hours per year – or bought power from a neighboring utility that had done so – to meet peak demands from customers. Electricity supplied from such marginal “peaker” plants is inevitably quite expensive, usually much more costly than the average price of electricity sold.
It was only in the 1980’s that utilities and regulators came to the realization it was both more cost-effective for customers and more profitable for utilities to install radio-controlled switches at customer premises to briefly turn off or turn down appliances, thereby modestly reducing electricity consumption during transitory peaks in demand. Thus, load management programs came into being.
These programs were generally effective and satisfactory to all parties. However, as standardized pre-set tariffs with non-negotiable terms, they were not price-based. In large part, this is because electricity market prices were not transparent – not only to the end customer, but even to the electric utility itself.
Only since the 1990’s have electricity prices become reasonably straightforward to obtain. In many regions (in the U.S. and in other countries), the electricity industry was restructured in several important respects, with the goal of creating more transparent markets and thereby introduce competitive forces to induce greater efficiencies in the sector.
As energy markets developed during the late 1990’s and early 2000’s, and competitors emerged to participate, certain new entrants discovered an entrepreneurial opportunity when prices became sufficiently high at peak moments: customers could be paid to use less energy at that instant, and the resulting “negative demand” could then be sold into the market as a supply resource.
Thus was demand response born.
Of course, to be accepted as a legitimate supply resource by wholesale power market administrators and operators, DR requires a firm measurable commitment by the customer to instantaneously reduce demand. Although turning down or turning off appliances is theoretically viable as a supply resource, in practice, transaction costs are prohibitive in verifying and aggregating demand reductions from many small-scale loads.
Consequently, the most common approach for DR has involved contracts with commercial or institutional customers in which standby diesel generators – installed in many buildings to maintain electricity service during grid outages – are turned on during peak periods. Not only is the resulting supply resource easily measurable, but also it usually comes in relatively sizable increments: at least hundreds of kilowatts, often at megawatt-scale.
Thus, portfolios of generator-based DR contracts were cost-effective to assemble and dispatch into regional power markets. When this was discovered in the first few years of this century, the DR markets quickly took off.
Why Would Anyone Think DR Is Dead?
Today, DR accounts for about 22 GW of supply resource in the U.S. While this represents only about 2% of the nation’s capacity base and peak demand levels, note that it’s the most critical 2%: the 2% that occurs at the moments of greatest market tightness and highest prices. In other words, DR is the marginal supply – and in economic markets, prices are typically set at the margin. At peak periods, electricity markets would be tighter and prices higher – perhaps much higher – without DR.
So, the advent of DR would seem to be a good story. And while presenters and attendees at the DR conference I recently attended were not deeply depressed, nevertheless the pervading sense was that the U.S. DR markets are not healthy.
Upon further investigation, the data does indicate DR market stagnation. Estimates from Navigant presented at the conference suggest that the U.S. DR markets have been essentially flat since 2011, and that 2015 represented a peak of volume that may not be surmounted until well after 2020.
Two primary factors have thwarted recent growth in the U.S. DR markets:
- Some of the most active regional wholesale marketmakers managing DR auctions – for instance, PJM and NYISO – changed some rules that made it more difficult for service providers to schedule or dispatch DR resources.
- Also, to reduce local emissions (especially in urban areas with air quality challenges), the U.S. EPA tightened restrictions on the use of standby diesel generators – the primary source of DR capacity that aggregators had been contracting in the marketplace from customers.
These dampening factors on growth were consequential. In 2017, many DR veterans were saddened, as two of the most notable pioneers of the DR marketplace were acquired: Comverge by Itron and EnerNOC by Enel.
Both Comverge and EnerNOC had ridden the early growth wave of DR in the early 2000’s and achieved very good exits through IPOs for their original investors. Such successes are rare in the energy venture arena, and these two former start-ups became accustomed to the spotlight as poster children for entrepreneurship, innovation and wealth-creation.
Alas, in recent years for the reasons discussed above, Comverge and EnerNOC had begun struggling to maintain profitable growth stories as independent entities. Each hit the wall, and were snapped up by larger players that sought to expand their list of services to retail customers of electricity. With that, the two primary superstars from the DR market no longer shined.
Reflecting upon these facts, it’s not hard to understand why a singular question was top-of-mind to a gathering of market practicioners: is DR dead, or dying?
DR Is Neither Dead Nor Dying, But It Is Changing
And so I found myself on a panel amongst DR thought-leaders confronting this existential question.
I confess now as I confessed then that I’m not an expert on DR markets. Indeed, I was invited to speak at ACI’s DR conference not about DR, but rather on the topic of behind-the-meter (BTM) storage involving energy storage devices located at a customer’s premises – one of the newer and hotter games in the energy services arena.
But, I felt strongly that the subject of my talk – energy storage – was indicative of the true answer to the question being posed. So, I threw it out for discussion:
No, DR is not dead or even dying. At worst, its position is secured and stable. However, I think the nature of DR is actually expanding at a fundamental level – provided that its definition is also expanded.
In reflecting further on the phrase “demand response”, I am struck by the customer-centricity of the concept. This orientation is relatively alien for the electricity industry, whose architecture was designed a century ago to provide electricity service to customers whatever their demands might be. In other words, the electricity sector has generally considered customers to be an exogenous factor, like weather, to be accommodated.
DR is a scheme to leverage the communal electricity grid – and its economics – to the potential benefit of particular customers on the grid. In ceasing to think about costs and instead focusing on customer value and market prices, it was a great leap forward for the electricity industry.
But now the electricity industry is boldly entering a new phase with dramatic consequences.
Rooftop solar is enabling millions of customers to produce electricity on-site, and sell surpluses back to the grid. And, BTM storage is allowing customers to store produced power or buy power from the grid at times of low prices either for later use or sale back to the grid at times of high prices.
In short, customers are becoming “prosumers” – sometimes producers, sometimes consumers – of energy.
Some industry observers think this will mean lots of customers “cutting the cord”, discontinuing electric utility service akin to the many who have gotten rid of wireline phone service and rely now solely on mobile phones.
While this may eventually become true for a few customers, I’m skeptical that many will ever actually leave the grid, for four reasons:
- Critically important to most customers, electricity quality can only decrease (perhaps substantially) by disconnecting from the grid – in ways that are much less satisfactory than “dropped calls” due to poor cell service.
- Unlike in telephony, there is no additional functionality and value (such as the ability to use intelligent apps at any time and place) that a grid-independent version of electricity service can provide.
- For many customers, it will be cost-prohibitive or otherwise impractical to acquire enough generation and storage capability to become completely self-sufficient in electricity – an irrelevant concern for telephone service.
- Unless and until wireless electricity transmission becomes viable (highly doubtful in my lifetime, and probably yours too), the full value of on-site energy assets – which are capital intensive – can only be monetized by transacting with other parties through the grid.
With the grid remaining a vital part of electricity supply for most customers, recall what DR truly represents: the ability for the customer to provide services (through reduced demand) to the grid at certain times for certain prices under certain conditions.
As more customers have on-site energy equipment that needs to be optimized relative to the grid, the spectrum of what looks (to the grid) like DR also increases.
In particular, from the standpoint of the wholesale market and bulk power system operator, BTM storage is essentially the same as backup standby generators. Thus, companies offering BTM storage services that include aggregation and dispatch of capacity back to the grid are already de facto offering DR.
If/when other forms of distributed generation (e.g., fuel cells) become economically viable, these too will need aggregation and dispatch into the grid and wholesale markets, just as is done with DR (and BTM storage) today.
Accordingly, I think that DR-type offerings are likely to in fact grow with greater penetration of customer-sited generation and storage assets. DR won’t just be limited to supply resources provided by price-based load control or backup generators.
So, DR is neither dead nor dying, just changing – albeit perhaps with a new semantic term warranted. This is not just my view: I was pleased to learn that my fellow panelists shared this general perspective. The discussion complete, we adjourned to mingle with an audience in somewhat better spirits.