Demand Response (DR) refers to mechanisms used to encourage consumers to reduce demand, thereby reducing the peak demand for electricity. Since electrical generation and transmission systems are generally sized to correspond to peak demand (plus margin for forecasting error and unforeseen events), lowering peak demand reduces overall plant and capital cost requirements.
The promise and hype surrounding the smart grid far exceeds its current capabilities. The commercial building sector, at which many DR programs are targeted, comprises 20 percent of energy demand and emissions in the United States. These concentrated pools of demand are a nightmare for strained utility grids, especially in major cities. As a result, utilities have created DR programs to provide financial incentives for building owners to reduce energy consumption during peak periods. But why would utilities pay end users to reduce power? The alternative is to build more power plants, which will cost the industry billions down the road.
The commercial building sector looks forward to the prospect of DR and the benefits that accompany it, namely reducing electrical demand during peak periods, which may result in monetary rewards for the customer.
Implementing Demand Response
As with anything, fundamental issues have limited and will continue to limit the success of these programs as they are rolled out. The most significant challenges include:
• The risk to tenant service levels posed by DR events;
• The execution complexities of intelligently managing through DR events, especially in large, complex buildings; and
• The capital requirements to fund the people or systems to ensure building owners and managers can address these issues.
Despite the potentially huge scale of DR programs, current technologies and systems that help utilities and the commercial building sector execute DR programs are surprisingly rudimentary. We'll call it "Manual DR." Some of the challenges with the current technologies are:
• Manual DR signals - emails, phone calls, text messages - from the utility or DR aggregator to a building manager to let them know a DR event is coming,
• Manual changes to energy systems (the building manager turning off lights or an HVAC chiller) to execute maximum load reduction,
• Limited ability to estimate load reduction quantity occurred versus target,
Limited ability to predict or determine effect on tenant comfort from the changes in the response, and
• Bill reconciliation and credit in 60-90 days when the utility collects and processes data.
How can building owners confidently participate in DR events this way? How can a utility build a scalable, effective DR program for thousands of customers? They can't. Commercial building owners and utilities say that the commercial sector has not been adopting DR programs to the expected extent. I recently spoke with one of the top five direct property investors in the world, and he noted that the company had only one building in the United States on a DR program, since that was the only facility with a building manager who knew the building and its systems well enough to manage an event. (Since that conversation, they have stopped participating.)
This lack of adoption is one of the biggest reasons utilities introduce a Peak Day Price (PDP) tariff program, which charges 10-15 times the normal rates during peak periods. Unlike voluntary incentive programs, large commercial customers are automatically moved onto the PDP program unless they opt out by adopting another DR plan.
Challenges to Realizing
So why don't DR programs work? Current commercial energy management systems and practices were not created for a world with DR (or for interaction with any external system, for that matter). The vast majority of recent industry innovation has come from "outside the meter" (outside the building) from groups such as EnerNOC or Comverge. In reality, these DR programs will not succeed unless there is innovation "inside the meter" (inside the building) to match it. Vendors such as Honeywell, Schneider Electric, and Johnson Controls cover "inside the meter" domains. Although these large companies have broad offerings, they have difficulty innovating quickly and don't have many incentives to radically change their mature, stable, and fairly concentrated industry.
To their credit, some of these companies are now focusing on solutions to these problems individually or through partnerships and acquisitions. Due to the work of government entities such as the Lawrence Berkeley National Laboratory, industry groups and some leading vendors are piloting more intelligent automated solutions and standards like OpenADR, a low-cost communications infrastructure to improve DR functions. But things are moving slowly. These emerging "AutoDR" solutions (which link facility energy management controls with external, utility-generated price or emergency signals) still leave much to be desired, as they consist only of an electronic signal from the utility or DR aggregator to the building management system (BMS), and a pre-programmed, fixed set of changes executed by the BMS over the building to respond to the DR event.
However, even "AutoDR" has its challenges. With these solutions, users cannot predict the actual load the building can tolerably shed, and they do not respond to weather conditions on the day of the DR event. There is also a limited ability to predict or determine the tenant comfort impact from response and limited ability to validate the load shed versus target in real-time. The results are the best efforts based on the fit between the pre-programmed response and the specific situation on the day of the event, which is close, but not exact.
In order to provide a scalable, widely-adopted, and valuable smart grid program for the commercial building sector, the industry needs a fundamental increase in the intelligence of their "inside the meter" DR systems, and they must also be tightly integrated with the "outside the meter" systems.
The next generation of DR will take "AutoDR" to the next level, and will include:
• An initial electronic signal from the utility or DR aggregator to the building manager and subsequent signals as the DR event progresses in the region with updates on any additional requirements.
• An automated response from the building's energy systems back to the utility or aggregator to confirm the precise capacity of load shedding available given current conditions.
• Incorporation of the DR event into the building's energy optimization plan for that day, with the energy plan tailored to comply with the event while minimizing tenant comfort impact.
• Dynamic adjustments to the energy plan and load shedding based on any changes to conditions (inside or outside the building) during the day.
• Estimated and real-time validation of actual load shedding versus target.
• Estimated and real-time understanding of tenant impact of response.
• Estimated and real-time understanding of program's economic benefits.
• Confirmation back to utility or DR aggregator immediately after the event confirming actual results.
If the industry is to overcome these barriers to DR adoption and achieve its smart grid potential, it must move from "Manual DR" and the emerging "AutoDR." It needs a range of solutions from the major "inside the meter" vendors to move the industry forward.