Demand Response & Electric Vehicles
By Mary Ann Piette
Demand response (DR) programs are developed and executed between electricity providers and customers to help change electric load shapes in response to economic or reliability issues with the electric grid. This is a topic of interest to large building owners because there are a growing array of programs available for owners to save money on electricity costs.
There are growing DR programs around the US and around the world. Figure 1 shows the DR capacity for commercial and industrial customers from a study to explore these developing markets. There is a large existing and technical potential for DR in 11 western states. These data are from the FERC National Assessment of DR with extensions from LBNL. The existing DR capabilities range from about 1% to nearly 11% of peak demand, with technical potential reaching 16%.
Figure 1. (a) Growing array of DR programs in the US, Europe, and in Asia and (b) a large savings potential in the Western US.
Applications
Historically DR has been targeted towards reducing customer demand during times of seasonal grid stress (typically hot summer afternoons or cold winter mornings or afternoons). Demand response strategies can be manual, semi-automated, or fully automated. Manual techniques require that a person actively participate in a DR event and change control systems for a given period of time, then restore them to normal. Semi-automated DR is a "man in the middle" configuration where there may be a set of automated responses programmed into a sequence of operations, with the trigger requiring a person to confirm or execute the command. Fully automated DR involves the electricity providers or DR program aggregator using an automated signaling system and the facility receiving signals. These signals are linked with pre-programmed control sequences in lighting or HVAC controls to automatically execute a change in the electric load shape.
The most common end-uses for demand response in commercial buildings are HVAC systems. There are a variety of strategies that can change the electric load shape depending on factors such as built-up or package unit systems or variable air volume or constant volume systems (Motegi et al, 2007; Piette et al, 2007.) Buildings with high thermal mass are good candidates for DR because the mass can be used to shift cooling loads and pre-cooling can be effective as part of a DR strategy. Many lighting systems can also provide demand response depending on how centralized the controls are. Dimming systems are capable of providing a fast and predictable load change.
Two key trends are motivating new DR program designs. One is that with improved automation and telemetry systems, building loads can provide fast ancillary services that are cleaner and potentially lower cost than conventional generators. Examples of the grid products that are developing into new markets are shown in Table 1. These products require DR systems that respond more often at different times of the day. Grid operators are looking for loads that can be reduced as well as increased. The second trend is that with the desire to provide greater levels of electricity from renewable sources such as wind and solar, the need for these advanced DR products are increasing. Developing an electric grid with more variable renewable supply is requiring the need for more flexible demand.
Table 1. Summary of new grid products for ancillary services (Kiliccote, et al. 2015).
An important technology trend in is the use of open standards for DR communication. All of the large California electric utilities are using Open Automated Demand Response (OpenADR) version 2.0 which is supported by the OpenADR Alliance. OpenADR 2.0 is also adopted by the Bonneville Power Administration. It is used in more than 8 countries around the world and in a growing number of utility programs across the US. In California the new building codes require that lighting and HVAC systems have the capability to receive and respond to a DR signal. This is a trend likely to continue in other parts of the US and around the world, and it is worth noting that all of the major building control companies are members of the OpenADR Alliance, offering many commercial products that are OpenADR 'ready'.
Figure 2 shows the client server architecture that is the core of OpenADR. Utility or grid operators provide signals to facilities using XML-based clients. There is a new language of DR signals in OpenADR that represent information such as integer electricity prices or grid signal modes.
Figure 2. OpenADR is a client server signals system where utility or grid operator's servers send signals to clients who are aggregators, campuses, buildings, or end-use system controls.
As DR programs evolve, dynamic pricing is key factor in DR and also a distinct issue in facility design and operations. The design of building, facility and systems needs to consider the electricity pricing available from the local electricity providers. There are many tariff designs around the US and around the world. There is a growing trend toward dynamic pricing. Most large campuses have to respond to time-of-use pricing and may have peak demand charges. Some regions have critical peak pricing that may include particularly high prices for 50 to 100 hours per year. The new advanced heating and cooling storage plant at Stanford University will be subject to day-ahead hourly prices. They will use a 7-day model predictive control to minimize energy costs (see discussions in District-level Energy Services and Diagnostics sections).
Another key factor in DR is energy storage. Large campus systems can often use storage systems to manage and flatten electric loads to minimize energy prices. Storage systems may consist of ice or chilled water, hot water, or electric batteries. The design and operation of these systems can reduce operating costs.
Another trend that facility managers need to plan for is the introduction of electric vehicles (EVs). These vehicles may include a local fleet that the campus manages and dispatches as well as EVs owned by the employees and visitors. EVs are a growing electric load that is plugged into the building or campus in the morning when employees arrive, and comes off the system when they depart. There may also be on-site electric fleets. The most important issue is the coordination of charging to ensure that electric loads are flattened and distributed.
Some innovative campuses are evaluating whether a fleet of electric vehicles can provide two-way, vehicle to grid storage. ( Marnay et. al 2013, Juul Frederick et. al 2015 )
Economics
The economics of demand response and storage systems is highly variable and related to the electricity pricing structures, incentives from utilities, and structure of DR program incentives. Some past customers have saved about 2% of annual utility costs by participating in DR programs. The first cost for enabling DR systems depend on whether the effort is manual, semi-automated or fully automated. Electric utilities in California are providing incentives of about $200/kW to $400/kW to pay for OpenADR.
With the growing EV loads at campuses and facilities the energy management staff needs to determine if charging will be free or if there is a recharge mechanism. There are a variety of techniques to manage charging infrastructures and commercial offerings.
Other considerations
A key co-benefit of demand response is energy efficiency. Many facilities managers find that when they participate in DR events there is an important feedback to basic energy management. For example, a common DR strategy is to set up zone temperatures for 4 hours. Many facility managers identify zones that are over cooling and may end up resetting normal operating conditions. Another example is duct static reset. A DR strategy that changes the duct static pressure for a short time can result changing normal set points. Another co-benefit of DR is load shape analysis and price response. Many facilities pay high peak demand charges, which can represent up to half of electric utility costs. Participating in DR events can provide information about how to modify electric load shapes to reduce utility bills.
A final area of co-benefits of DR is the consideration of the capability of the system to run in low-power modes. Future electric systems will have more distributed generation, storage, and features common to emerging micro-grids. Having distributed DR capabilities allows a local system to run in low-power mode during emergencies and grid outages. A DR capable system has the ability to run on reduce power mode for sustained periods. An advanced DR system may be able to reduce lighting, HVAC data center, and plug loads.
The development of the ability to reduce electric loads for transactions with the electric grid is an important capability that is emerging in energy management. The need and nature of these transactions will continue to evolve over the next decade as we have more distributed generation and lower cost control automation. These capabilities have synergistic benefits with daily energy management and are valuable as electric prices become more dynamic and time differentiated.
Institutional requirements & capacity
There are training needs associated with demand response since it is a new area of operations. Facility managers need to understand the economics of participating in a program as well as the control strategies the facility plans to execute. In addition, commissioning practices will need to be modernized to address DR systems and protocols.
Similarly there are new electrical system requirements for EV charging stations. Facility managers need to evaluate the location and use of these stations to provide new services to the growing number of EVs at their sites. Some facilities instigate recharge systems to the local staff or visitors.
As more EVs are located on a facility's electric meter, the dynamic charging system can use the EV load as part of a DR strategy and defer the charging for a few hours, or cycle the cars to manage the peak demand. The basic concept is to require that when a car is plugged into the charging system, the requirements for how fast it has to be charged are recorded and the charging is managed as part of the total load management activity. If charging is no needed to be completed until the end of the work day, the charging load can be flattened and managed as part of the flexible load for cost minimization.
References
Federal Energy Regulatory Commission (FERC), National Assessment and Action Plan on Demand Response. 2009. A National Assessment of Demand Response Potential.
Juul Frederick, Negrete-Pincetic M, MacDonald J, and Callaway D. 2015. Real-time Scheduling of Electric Vehicles for Ancillary Services, IEEE Power & Energy Society General Meeting 2015.
Kiliccote, S., D. Olsen, M. Sohn, and M.A. Piette. 2015. Characterization of Demand Response from Commercial, Industrial and Residential Sectors. Accepted for publication in Wiley Interdisciplinary Reviews: Energy and Environment. Edited by Peter Lund.
Marnay, Chris, Chan TW, DeForest N, Lai J, MacDonald J, Stadler M, Erdmann T, Hoheisel A, Mueller M, Sabre S, et al. 2013. Los Angeles Air Force Base Vehicle to Grid Pilot Project. In ECEEE 2013 Summer Study on Energy Efficiency. LBNL Report LBNL-6154E.
Motegi, NA., M.A. Piette, D.S. Watson, S. Kiliccote, P Xu. 2007. Introduction to Commercial Building Control Strategies and Techniques for Demand Response. Report for the California Energy Commission, Lawrence Berkeley National Laboratory. LBNL-59975.
Navigant Research. 2015. Demand Response Enabling Technologies: Metering, Communications, and Controls Equipment: Global Market Analysis and Forecasts.
Piette, M.A., D.S. Watson, N.A. Motegi, S. Kiliccote, P. Xu. 2007. Automated Critical Peak Pricing Field Tests: 2007. 2006 Pilot Program Description and Results, LBNL 62218. PG&E Emerging Technologies Program.
More Information
OpenADR Alliance web site - http://www.openadr.org/
Vehicle to grid simulation tools - http://v2gsim.lbl.gov/