November-December 2009

Tracking Energy Usage

Submetering offers the ideal strategy for peak shaving and managing demand.

The growing trend of tracking energy usage with submeters is reducing demand in both industrial and commercial settings. Moreover, it’s finding favor as a key to unlocking maximum performance in peak shaving applications. In fact, submetering data is a critical component and should be the first step in any peak shaving strategy, according to Lindsay Audin, President of Energywiz Inc., in Hudson, NY.

Audin has consulted on submetering and energy issues with a broad range of industries, and teaches the art of analyzing submetering data at Association of Electrical Engineers–sponsored seminars, such as “3D Load Profiling Using Interval Meter Data.” It’s the 3D (three-dimensional) aspect that allows managers to understand the impact of energy usage and how to tailor the ideal peak shaving strategy.

“I’ve been teaching this for five to 10 years, and people are always amazed,” says Audin. “When you look at these three-dimensional profiles, you can see a continuous service [record] that looks like a mountain range with flat valleys, plains, peaks, and plateaus, and those are very demonstrative as to how the electric load varies.”

Audin notes that when the information is presented in a graphical format, it’s easy to find and correlate the load demand with factors such as weather, occupancy, equipment, and facility usage. Such an exercise recently saved a hospital from installing a distributed energy system that was far larger than the requirements of their load demand.

“We looked at their existing hospital because essentially they are moving everything from an old building to a new building, which is substantially larger,” says Audin. “The submetering data showed that the cogeneration system they were proposing was roughly double the size it needed to be. So now it’s been scaled back in size, both on the thermal and electrical load, and it will be much more cost-effective.”

Audin says that energy is typically the third-highest cost at hospitals and research universities, exceeded only by mortgage and labor. If we’re talking about a research university in an area with peak rate billing, such as the University of California at San Diego (UCSD), peak shaving with distributed energy is an effective solution. But getting the highest performance from the system means controlling demand, says John Dilliott, facilities engineer at UCSD.

The UCSD campus has a microgrid with a load of roughly 40 MW. A combined heat and power plant driven by a 30-MW gas turbine, plus a 1-MW photovoltaic solar array, allows the campus to save money when the local utility’s costly peak rates are in effect, but lowering the demand can save even more.

“Our cogeneration plant has a flat cost for energy, while the imported electricity is a time-of-use issue,” explains Dilliott. “So we participate in demand response programs and try to manage our peak. One thing we want to do with the submetering is get feedback to the occupants, showing them where the peaks are and when they need to reduce energy usage.” 

The university started a submetering program in 2003 to raise energy efficiency and apply for State of California incentives. “We wanted to find the buildings that were the bad actors,” says Dilliott. “We had a general idea, but digital meters show demand every 15 minutes, and we wanted data on an intensity basis instead of a total usage basis.”

Rather than make the mistake of assuming that a building’s size should be the sole determinate in focusing the program’s energy efficiency efforts, the team looked at energy use per square foot, and the results determined the initial request for rebates to fund energy efficiency projects in those buildings. An additional benefit of the effort is the fact that it can build awareness of energy costs for a building’s occupants.

“Building occupants aren’t billed for their electricity,” says Dilliott. “The data will serve as an incentive for voluntary conservation efforts.”

The final step in the program will evolve with the university’s vision of a smart grid. “We’re looking to increase our submetering capability and get more granular information,” says Dilliott. “Without having detailed usage profiles, it’s not an efficient smart grid, because you want to have smart buildings. When we look at demand on a 15-minute basis, we can tie that in with management and demand response.

“We will be reducing our dependence upon the grid and taking more advantage of those 30 megawatts of cogeneration,” he continues. “The bottom line is reducing the imported power and the money we pay for it.”

Reducing power was a priority of the US Postal Service in the Greensboro, NC, when it implemented submetering control and monitoring services from ADMMicro Inc., Lynchburg, VA. At the Post Office’s Greensboro Guilford location, a 6,000–square foot facility, energy usage dropped by 31%. Altogether, the total savings for 16 facilities was $52,000 over a 13-month period. The key factor was the ability to automatically shut off HVAC service to unoccupied facilities and return them to normal just before employees returned to work.

“You have the ability to look at how energy flows down to major loads such as compressors, data racks, or HVAC systems,” says Don Howell, CEO of ADMMicro. “Once a customer understands their usage, it opens up opportunities for peak shaving and other activities.”

So, how does it work? Take a facility where the data reflects real-time energy utilization at the device level. By creating load profiles, a monitoring and control system such as ADMMicro’s can make minute changes in the devices it controls and prevent the total load from increasing. For example, although a facility manager could manually shut down a rooftop HVAC unit, leaving it off for any length of time might have an adverse impact. A more desirable method would automatically curtail one stage of that unit and lower demand within a certain window of time.

If the real-time monitoring revealed that there wasn’t enough curtailment, the system could shut down another stage of the HVAC, or curtail a different load, such as dimming lights. “It may be just 2–3%,” says Howell. “That’s much different from shutting off a whole bank of lights. These changes are done in a very finite manner by making minor adjustments.

“The cumulative effect on the demand is surprising,” he says. “It’s along the same concept as peak shaving and similar to when the utility has time of demand penalties and the company wants to curtail its load to avoid those.”

And demand penalties are certainly on the rise. Utilities across the nation are moving towards interval meters, and in New York City, NY, Con Edison has installed an interval or “smart meter” at any location with a peak demand of more than 1,500 kW. Recently, Con Edison announced that it is dropping the limit from 1,500 kW to 500 kW, after November 2010. In New York City, the typical office building could experience a 40% or 50% higher bill due to peak demand charges, according to Audin.

“Now you're going to see a much bigger differentiation in the cost of electricity among customers based upon their usage patterns,” says Audin. “Those who are sloppy and leave everything running at two or three o’clock in the afternoon are going to see their bills skyrocket. Those who are able to shift their loads are going to see a bonanza.”

The good news is that Con Edison customers may get some help in acquiring submetering technology.

New York City–based CPower, an energy management company using submetering and other methods, has an aggregate peak load management contract with the New York State Energy Research and Development Agency, for $8.9 million in incentives. The contract targets various peak load reduction projects to attain more than 14 MW of peak load reduction in New York City and the surrounding areas. Projects eligible for peak load reduction include: load shifting or load curtailment, clean emergency generation for load shifting, cogeneration, and distributed generation.

Obviously, companies will save the most with peak shaving strategies when they know the exact size and schedule of their loads. Moreover, combining a distributed energy solution with submetering data can offer the most efficient and cost-effective alternative to expensive utility-supplied power.

How to Find the Right Submetering Solution
Saving money with submetering starts with a brief talk with a utility account representative. Companies need to understand existing utility rate structures and how alternative rate structures can justify a greater analysis of electricity usage. In the case of voluntary time-of-day rates, this form of pricing is a double-edged sword that can expose a customer to a greater level of risk, but also a greater level of opportunity.

“Go to your account rep and make sure you understand how you are paying for electricity and your options,” advises Audin. “Then, talk to an energy consultant or meter data service provider. In essence, there are various levels of expertise and vendors. It all depends on your needs, your electric rates, and your opportunities.”

The good news is that wireless technology has simplified submetering. Wireless submeters typically communicate with a central processing system through cell phone frequencies, and they can monitor loads without cutting into wiring, because they can sense the amount of current traveling through a cable.

“It looks like a doughnut that clamps around a wire,” says Audin. “You put these temporary data loggers in a variety of places and survey your usage. The loads are monitored for about a month, and charts are created that help you to understand your individual loads.”

For real-time access to data, submeters can be monitored through an Internet connection.