Considering the loud din emerging from suppliers, utilities and industry advocates for Demand Response, it would be helpful for prospective participants to have an understanding of what is involved in these programs as well as what is required to be successful in the implementation of a Demand Response program. Demand Response in its most basic form is the commitment by a building operator to reduce the electrical demand of one or more facilities in response to a request by the servicing utility or by market based pricing. Utilities want large consumers (or many small consumers) to be able to rapidly and effectively reduce electrical consumption in response to an electric power supply (or gas supply) shortage or an exceptionally high demand on the power grid or simply by market pricing. Generally, large commercial buildings and many industrial plants are prime candidates for Demand Response; however, schools, hospitals, retailers and public and many small commercial buildings may also be candidates.

[an error occurred while processing this directive]

Demand response may be a formal program pursuant to an agreement with the servicing utility that provides incentives for participants who commit to load shedding on demand. Demand Response may also be an informal program where facility managers want to be a good “corporate citizens” and reduce demand during periods where power is in short supply. In either case, effective demand response requires knowledge of typical and individual building performance which implies a further knowledge of performance capabilities vs. occupant comfort and business support. This article assumes that Demand Response is to be implemented and will cover many of the considerations and pitfalls that facility and energy managers need to address for successful execution of a Demand Response program.

What are the costs and benefits from a Demand Reduction program and what factors need to be considered in the implementation of such a program? While costs and benefits are beyond the scope of this article and are specific to each situation, a look at the other implementation factors is certainly worthwhile:

• What is the “typical” demand profile for the facility during the expected time of a Demand Reduction event? Having a reliable “baseline” from which to calculate expected results and measure actual demand curtailment is a necessity.

• What is the practical demand level that the facility is capable of achieving after load reduction? Many facilities (with the possible exception of offices and manufacturing facilities) can reduce lighting loads with a minor impact on business. HVAC can be generally reduced with a minor impact on occupant comfort. Refrigeration loads can be temporarily reduced where it does not impact food quality (e.g. beverage coolers).

• Does management have the infrastructure to implement, monitor, and manage Demand Response events? How is Demand Response event initiated and communicated to systems and personnel involved or impacted by the event?

• How is demand reduction to be measured and success of the program to be determined? Facilities with significant demand fluctuation and varying business operations are more difficult cases in which to specify and measure a demand shedding level.

• How much are business operations and occupant comfort affected and does the reduction strategy have ways to mitigate this if there is a problem?

• Are generators to be started to further contribute to the reduction of load on the grid? If this is the case then the generator must be monitored and controlled and the cost of generator operation figured into Demand Reduction economics.

Determining the Normal Facility Demand Profile
Before committing to a load reduction for a facility, it is recommended that the facility be in reasonably good working order. This does not necessarily mean “re-commissioning” the facility but all systems should be functioning normally. The baseline should be established over a period sufficiently long to allow for the daily variances that effect building load (e.g. outside temperature, building occupancy, deliveries, equipment maintenance, etc.). Any large increase or drop in demand that is seen in demand profiles should be understood and a determination made as to its significance. Statistical as well as graphical analysis will help establish a baseline.

If HVAC is the primary source of loads to be shed, then regression methods can also be used by more sophisticated operators to account for occupancy levels and outside temperature which directly and markedly effect HVAC power usage. These can render “adjustment” factors for facility load profiles to account for outside air temperature. Calculating occupancy levels is usually not practical, but can generally be considered to be a function of the day of the week and time of day, particularly in retail commercial facilities. Comparing a “Friday” demand profile to another “Friday” demand profile is a more valid comparison than comparing the same “Friday” profile to a “Monday” profile.

[an error occurred while processing this directive] Determining the Amount of Demand Reduction
After committing to participate in a Demand Reduction program, the facility/energy manager should determine the amount of demand that can be cut from a facility or the maximum demand to be permitted at the facility depending on how the reduction is to be measured. The former method requires that sufficient loads be shed to achieve a predetermined demand reduction. The latter method requires the facility not to exceed a prescribed total demand level for the duration of the event.

In either case, the reduction goal, however specified, is achieved by shutting off non-essential electrical loads until the performance goals are met. Loads may be curtailed by merely being turned off or they may be curtailed as a result of modifying setpoints to alter normal operations. In the latter case, the controlling software must have the additional capability of assuring that all units do not turn off and then all come on which would effectively negate demand control. It is recommended that an analysis be performed to determine what loads are available for curtailment and to what degree they can be curtailed. Involving the building occupants/operators in this may be a good idea. People want to “do their part” in conserving energy but too much discomfort for employees or customers can hurt business or may even compromise safety.

Loads to be curtailed are determined by a number of factors:

• Size of a load – Large A/C units produce more results (“bang for the buck”) for a single control point than small units. Controlling lighting produces consistent demand reduction; however, lighting reduction may be dependent on the availability of sunlight during the event. Controlling HVAC, particularly on moderate temperature days, produces inconsistent results. Hot water can be usually be controlled but not during washing periods for food preparation areas.

• Ease of controlling the load – How much does it cost to control the load and how is the load to be “expertly” controlled during an event. Loads controlled by an installed energy management system/building control system can generally be easily and expertly controlled by modifying the programmed control routines. Modulation of controlled loads (e.g. slowing variable speed fans, chillers, pumps, etc.) if available, is generally preferable to on/off control.

• Importance of the load to the business – What is the impact on the business of shutting a load off or raising the temperature of an interior conditioned space?

After evaluation of the above criteria and some testing, the demand shedding process for a facility can be reduced set of operating rules and parameters for the building control system that will be implemented at the beginning of a Demand Response event. These control rules are programmed into the building/energy management system for each facility and become an integral part of the enterprise program. An individual facility Demand Response scheme should, as a minimum, contain the following:

• Total load available for curtailment (can be calculated from HVAC operating data, lighting data, etc.)

• Target load to be reduced (curtailment of all available loads for the entire duration of the event is probably not practical nor advisable). Consideration should be given to curtailing a percentage of the available load subject to curtailment according to the “urgency” of the demand reduction.

• Ranking (i.e. importance) of the loads available for curtailment (some building loads may not as “sensitive” to curtailment as others) to determine the order of loads to be shed to achieve reduction goals. This facet of the Demand Response scheme can be as simple as shutting off (or the coordinated duty cycling) of loads identified for curtailment or it may involve a more sophisticated plan to shed loads less indiscriminately. Consider identifying loads to be shed according to the present impact on occupants and operations. A discussion of this concept follows in this section.

In most, if not all, facilities subject to Demand Response, the facility operator seeks to achieve the required/programmed amount of load to be shed and do so with the least impact on the business or services housed in the facility. This should lead the Demand Response program manager to working with individual facility manager(s) to determine the order/priority for shedding the loads. The order for shedding may be determined by the present impact of the load curtailment on occupants and operations. For instance, if a HVAC zone is unoccupied, then the impact is non-existent. If a zone has a large number of customers present then the impact may be significant. Consider constructing a matrix whereby loads are shed dynamically by their present impact which may be represented by a factor that considers the following:

• Perceived/ranked importance of the load to the business. Rank HVAC zones by criticality to business operations (Conference Rooms may be high and Shipping and Receiving may be low) and the value to the “setback” temperature that can be implemented during a Demand Response event. Lighting can similarly be ranked by use and amount of reduction that can be tolerated. Shed low priority zones first.

• How close the load is to a tolerable operating point. Consider writing new setpoints to HVAC zones during a Demand Response event and shutting off units that meet or exceed a new setpoint. Consider shedding lighting according to reduced minimum light level criteria and particularly if a lighting area is unoccupied.

• At set intervals during the shed event, reevaluate loads that are shed. It may be necessary to turn on some loads if zone temperatures are excessive or light levels are not practical. Turn off other loads that have been running according to established criteria.

• Consider using user/occupant input to adjust priorities. Local conditions may cause a given load to have a higher priority during a particular Demand Response event than it would usually have.

Determining the actual amount of demand to be reported as being shed during a Demand Response event is, in its simplest form, the mathematical difference between the actual average demand of the facility during the event and the average baseline demand for the facility during the same time period. A more accurate assessment of the demand reduction would require mathematical integration of the average demand level of the facility during the event at 1-15 minute time intervals and compare this with the integration of the facility baseline profile during the same period. Other methods that account for variations in occupancy levels and outside temperature would provide an even more accurate assessment for larger, geographically diverse operations.

It is recommended that an assessment be conducted on the non-quantitative aspects of Demand Response events. After a lighting reduction, is the facility still functional or do customers leave the building. Are employees still productive in a warmer building? Did customers complain? Often some public relations efforts to communicate the importance of Demand Reduction to both customers and employees goes a long way towards sustained occupant satisfaction and support of the program.

[an error occurred while processing this directive] Infrastructure Needed to Support Demand Reduction
The infrastructure required to support a reliable Demand Response program should be determined after a detailed plan is adopted for the program. The plan, which includes the requesting utility and extends though the enterprise facility manager to the individual facility, specifies in detail all aspects of the Demand Response program.

• Demand Reduction requirements/expectations

• Procedures for initiation and termination of events

• Confirmation of notification receipt and response by facilities

• Confirmation of termination and resumption of normal operation at facilities

• Agreement on expected results of the Demand Response program and how results are measured

• Measurement of results and determination of both the value to the utility and the value to the enterprise

Once the plan is in place, then the facility manager should survey available resources (i.e. manpower, IT, EMS, analytical/accounting, etc.) that will be needed to implement and operate the program. If resources are not available, then plan modifications will be needed. A Demand Response program will not “run itself” and will require considerable resources at all levels to be successfully implemented. Ongoing manpower and IT resources will be required to maintain the Demand Response system, evaluate its operation, and implement needed changes.

After surveying available resources, the program manager should begin applying these resources against program requirements in what will become the basis for a plan of implementation and specific tasking. Aside from the needed approvals and “buy-in” characteristic of implementing a program within a large organization, the program manager will then need to manage the plan by ensuring that all required Demand Response requests are validated and processed efficiently. The plan/program should reflect the basic elements of a Demand Response mechanism as depicted in the flowchart below:

The effective enterprise Demand Response mechanism should thus consist of program elements from the utility/ISO, enterprise facility/energy management, and each participating facility. These program elements should be supported by personnel, software, Information Technology (IT) assets, and, ultimately, by facility monitoring and control systems. Ultimately, software, monitored by responsible managers can be expected to accomplish most program tasks.

In normal operations, a demand reduction request from the ISO is accepted, understood, and acknowledged by the enterprise. This request is then translated into a series of event specifications which should include, as a minimum, the participating facilities, demand reduction amount/demand limit required of each facility, the start date/time of the event, and the termination date/time of the event. Next, individual Demand Response messages are transmitted to each involved facility. Each facility, in turn acknowledges receipt of the message. Facilities should then confirm implementation of Demand Response actions and calculate demand or demand reduction levels at specified intervals. Finally, each facility should confirm termination of the Demand Response event and return to normal operation. Data from each facility should be captured in the enterprise database which also contains Demand Response program specifications in order for the success of the Demand Response program to be measured and reported.

From the above, participating enterprises and facilities should have active, real-time control of devices responsible for significant electric (or, potentially, gas) consumption. For facility operators with multiple facilities, this capability should be read as some measure of centralized control and monitoring of facilities. This capability allows the Demand Reduction participant to effectively manage the program and to ultimately mitigate exceptionally high demands for power. In return, the utility or ISO can provide some level of power to users. In simple terms, better to have some power than no power.

Effective implementation of Demand Response at an individual site, and in particular a large site, is best accomplished through an intelligent site energy/building control system. Many Demand Response implementations consist of an electric meter with a communicating relay used to initiate and terminate the demand response event through a local control system. This solution is marginally effective and seldom provides sustainable results. A more effective solution requires network connectivity to the existing building controls and metering systems and supporting enterprise software. This solution, which includes feedback and evaluation mechanisms, allows for maximizing load shedding at the facility while minimizing the impact on the facility. By closely monitoring the operation of the facility during the Demand Response event, performance of the program can be monitored and enhanced over time.

Once the Demand Response mechanism is fully functional then the energy manager may want to consider program enhancements. Such enhancements may include the continuing evaluation of facility energy performance to identify under-performing facilities and apply needed corrections. The enterprise may want the ability to internally initiate limited demand reductions in response to real time energy pricing fluctuations. Further, enterprises see the advantage of hedging energy costs by predicting energy usage and adjusting energy procurement against a controllable demand. These program enhancements can be implemented with today’s technology if they are supported with a comprehensive enterprise energy reporting system.

[an error occurred while processing this directive] Measuring Demand Reduction and Program Success
The success of the Demand Response program, as has been pointed out in most of the preceding text, is contingent on accurately measuring and recording the avoided demand on the electric grid contributed by each participating facility. This is critical to not only evaluate the operation of the program as a whole, but to optimize and maintain individual facility performance in response to changes in building operations, outside conditions, and system maintenance. In addition, the facility manager must constantly monitor and record changes to building systems and operations in addition to ongoing building energy performance. This may lead to adjustments in the building energy baseline or in the expected demand reduction that the facility is expected to contribute.

Measurement of the efficacy of a Demand Response event is measured at the main facility meter – the total load place on the grid by the facility. This obviously requires that the facility report, in real time and at specified intervals, its demand both during normal operations and particularly during a Demand Response event. It does little good to learn that there was a problem at a facility yesterday and the demand reduction opportunity was lost. Besides potentially being costly, this erodes confidence in the program. In addition to data from the main building meter, additional data from connected monitoring and control systems (e.g. zone temperatures, equipment status, light levels, outside temperature, occupancy, etc.) as well as any sub-meters that may be installed, provides valuable information that allows for fine tuning and troubleshooting Demand Response actions at the site level.

The enterprise must be able to acquire, consolidate, compile, and analyze available facility data to support an effective Demand Response program.

Generators
Generators are an asset that are sometimes included in a demand reduction program, particularly when a facility in the program has one or more sizeable generators. Hospitals are one type of facility that usually has a sizeable emergency generator; however, many retail and critical office facilities are also installing generators, particularly in disaster-prone areas. Inclusion of generators in a Demand Response plan requires careful analysis by the facility manager of all of the factors involved:

• Net generator hourly operating cost to include fuel and maintenance

• Regulatory restrictions (Hours or conditions of operation may be restricted.)

• Servicing (fuelling, maintenance) capabilities

• Remote dispatch and monitoring capability (Don’t depend on local personnel wanting to start, stop, and monitor the generator for a dispatch event.)

Consideration of each of the above in consideration of the value of power that a generator can produce during a Demand Response event will ultimately guide the decision to utilize generators to mitigate the demand target for a facility.

Summary
The design and execution of an effective Demand Response program requires a methodical analysis and deployment of many layers of systems. Sustainable, long-term load shedding capability must be supported by sound facility and energy management expertise reinforced with sufficient data to confirm compliance with the program and to identify any shortcomings. A system of measurement and comparison, checks and balances, will insure that the intended plan is actually executed and delivers the expected results.

The inclusion of facility occupants and customers in the Demand Response plan and the measurement of its success will help ensure the viability of the program. The successful facility will be one where the occupants are aware of the importance of the program, are informed of the occurrence of a Demand Response event, and are given, to the maximum extent practical, some means to mitigate the effect of an event on building operations and occupant comfort. Eliminating grassroots resistance to its implementation and use early on and addressing operator concerns in a timely manner will enable the facility/energy manager to concentrate on many of the other aspects of managing the program.

About the Authors

David Wolins is Vice President of EnFlex Corp. As a founding member of EnFlex, David brings 17+ years of energy consulting in HVAC, refrigeration and controls. He has worked with retail, commercial and institutional clients world wide.

About EnFlex
EnFlex markets an open and totally integrated facility information system consisting of a suite of hardware, software, and support services that deliver scalable facility monitoring and control solutions. EnFlex specializes in providing connectivity and web-enabling solutions to effectively integrate the existing facility monitoring and control systems of its customers. EnFlex supplies services to customers to assist them in analyzing facility operations and implementing energy, demand, and controls optimization programs.

Paul Hepperla is Director of Energy Services at Verisae, Inc. Paul brings to Verisae more than 13 year’s experience working with retail and commercial clients and extensive knowledge of the energy industry. Prior to joining Verisae, Hepperla was Manager of Energy and Operations Services with SUPERVALU, the 2nd largest grocery retailer in the United States. He has also served in various roles with Xcel Energy, Johnson Controls, Inc. and Building Automation Products, Inc. Paul has a Bachelors of Science in Aerospace Engineering from St. Louis University.

About Verisae
Verisae (www.verisae.com) is a leading supplier of maintenance process management and asset management software and services with offices in the US, UK, and Asia. Verisae’s enterprise asset and maintenance management software improves work flow and helps retailers reduce operating costs and meet refrigerant compliance requirements. Products include asset management, maintenance management, service automation, and electronic payment of service providers, enterprise asset procurement, and enterprise energy management. Verisae provides services including site asset surveys, refrigerant inventory surveys, and consulting to assure an effective implementation of the solutions

[an error occurred while processing this directive]
[Click Banner To Learn More]

[Home Page]  [The Automator]  [About]  [Subscribe ]  [Contact Us]