Princeton University’s Ted Borer offers historical perspective on lighting — going back to before Thomas Edison — and explains why CHP microgrids offer so much value to today’s electric grid.
Electricity is a high grade form of energy. It is clean to use, high energy-density, quiet, and has a vast array of applications. There is a well-developed distribution infrastructure and the properties of electricity are well understood. Electricity can be generated by conversion from many different sources including: hydro, nuclear, solar, wind, tidal, coal, oil, gas, biomass, waste… Electricity has become a “common currency” for energy.
We are accustomed to reliable, consistent electric power in almost every building we enter. We take it for granted that with the flick of a switch we can enjoy lights, heat, cooling, and an infinite array of modern conveniences. But, everything that uses energy requires maintenance. And at some point, each thing we build will fail. The important question is not whether the components of electric power systems will fail, but how well we have designed the system so that failures cause the least disruption, and how well prepared we are to restore them to normal service.
Advantages of local storage, production, and control
Grid-scale reliability was not always a concern. Before electric lighting or gas lamps, kerosene lamps included local fuel storage and direct control by the user. By having a diverse array of non-connected lamps, there was little risk that a single lamp running out of fuel would result in failure of all lighting throughout a building, let alone, a town. Even if a building ran out of fuel, there was likely to be some stored next door. There were many drawbacks to the use of kerosene lamps, but regional common-mode failure was not one of them.
In the late 1800s, electric illuminating companies were developed. They offered advantages including: safety, reduced operating cost, convenience, diverse fuel and generating assets, reliability, environmental impact located far from the user, and a predictable and measurable product. But with those benefits came more potential failure points along the path between fuel supply, and the end-user’s lamp. The end user had lost full control. Some points in large power systems could fail and cause many users to lose service concurrently. Obviously, modern power grids are designed to provide multiple supply paths wherever possible and cost-effective. But even with built-in redundancy, weak points necessarily exist throughout all power distribution systems.
Increasing the scale of an electrical distribution system may improve efficiency, but it increases the risk of regional blackouts
It is more cost-effective to transport electricity over long distances than to transport the fuel and water used to make that electricity. Utility plants were generally built closer to fuel and water resources and far from end users. The aesthetic and environmental impacts of power plants were also distant and hidden from most users. Centralized power plants could be located on lower-value real estate and were designed to benefit from efficiencies of scale. In the 1880s, Thomas Edison’s Pearl Street Generating Station in New York City included six electric “dynamos” with a capacity of 100 kW each. Today’s utility generators can exceed a million kW each. There are often multiple large generators on a utility site. They are more efficient than Edison’s, but entire regions can be affected when a modern scale generating system fails.
To maintain reliable service with consistent voltage and frequency, modern power grids include many generators at diverse locations. The North American Electric Reliability Council also requires grid operators to run some generators at less than their full capacity. These can be called on to rapidly increase output in the event of a failure of a different generator or spike in power demand. These generators are paid to run as “spinning reserve.” The risk of loss of any one generator is mitigated by having enough generators running at partial output that the loss can be absorbed by the spinning reserve capacity. At a minimum this concept requires spinning reserve capacity to equal or exceed “N-1” criteria, i.e., enough spinning reserve to replace the loss of the largest generator on the grid. The largest generating systems in the Pennsylvania-Jersey-Maryland Interconnection (PJM) grid deliver about 2700 MW each. So PJM maintains at least 2700 MW of spinning reserve. This capacity uses fuel while producing no power. Thus, reserve capacity is zero-percent efficient. On any given day spinning reserve might represent 1 percent of total PJM capacity. Using 1 percent additional energy may be a small price for added reliability, but even that amount need not be wasted with careful grid design.
So a grid operator is faced with competing desires of having plants as large as possible for efficiency of scale, but plants that are as small and disparate as possible to diversify the risk and magnitude of failure. As the amount of spinning reserve is increased to add reliability, the net efficiency of the power grid drops. Here, low operating cost and high reliability are in opposition.
Today’s typical power grid efficiency
On average, about a third of the energy purchased as fuel is actually delivered as electricity. Where does the other energy go? It is lost as heat. In a typical utility generating plant, about two thirds of the original fuel energy is rejected to the environment as heat through a chimney, cooling tower, or condenser that dumps heat to a river, lake, or ocean. The temperature of this waste heat is not high enough to be cost-effectively converted to electricity.
Couldn’t the heat from power generation be used productively by someone?
Yes. “Waste” heat still has value. Unfortunately, it is less cost-effective to transmit thermal energy over long distances than it is to transmit electricity. Utility plants’ residual heat has much lower “energy density” than electricity. The piping networks used to deliver heat are costly to install. Utility plants located far from urban centers suffer from the disadvantage that they don’t have nearby customers who can make good use of the thermal energy that is a byproduct of power generation. To enjoy efficiencies higher than today’s grid, the “waste” heat from power generation needs to be produced near a community that values it.
But there are many communities that value heat and operate district heating systems. The U.S. Naval Academy in Annapolis began district steam service in 1853. The first commercially successful district heating system was launched in Lockport, N.Y., in 1877. Philadelphia and New York City both have extensive district heating systems. District energy provides heat for most of the federal buildings in Washington, DC. Denver’s district steam system, built in 1880, continues service today. District heating systems are operated in: Boston, Cambridge, Detroit, Harrisburg, Lansing, Minneapolis, Omaha, Pittsburgh, San Diego, St Paul, San Francisco, Seattle, Trenton, and dozens of other cities around the country. Colleges using district heating include: Case Western Reserve, Cornell, Iowa State, University of Maryland, MIT, Michigan State, The University of New Hampshire, North Dakota State, University of Notre Dame, Princeton, Purdue, and University of Texas at Austin.
Some of these district energy systems produce not just heat, but combined heat and power (CHP). They enjoy lower operating costs and lower life-cycle costs than if they produced just heat and purchased power separately. It is common for CHP systems to operate at efficiencies of 80 percent and more. The life-cycle energy savings far outweigh the initial cost of infrastructure. By reducing total fuel used to produce heat and power in combination, cost as well as CO2 and other emissions are reduced proportionately. Electric power systems that normally operate synchronized with the regional power grid, but that can separate and run autonomously in an emergency can be considered microgrids.
By operating a district energy plant, communities with CHP enjoy lower total energy costs and lower labor requirements. They have centralized fuel deliveries, and maintenance. Less space in their buildings is occupied by heating equipment and more space is available for programming. Most of these systems were built to save money. But they also enjoy higher power reliability in an emergency.
The same communities who sought to save money and lower their carbon footprint can also maintain service to critical electric customers during emergency conditions. By normally operating their generators synchronized with the power grid they take advantage of the grid’s usual reliability. But they are able to operate as islands when the grid fails. It is no coincidence that the following CHP microgrid systems all rode through Hurricane Sandy with the heat and lights operating, even while the surrounding towns were dark:
- Co-Op City, Bronx NY, 45 MW CHP
- Princeton University, 15 MW CHP
- New York University, 10-MW CHP plant
- Pepco Thermal Energy Plant- Atlantic City, NJ, 5.7 MW CHP
- Danbury Hospital, CT 4.5 MW CHP
- Hunterdon Developmental Center Clinton, NJ, 4.5 MW CHP
- Hartford Hospital, CHP
- Fairfield University, CHP
- Nassau County Cogen (supports hospital)
By remaining in operation, they not only maintained their own mission-critical activities, but also could be points of refuge, supporting the surrounding communities.
As a result they now enjoy lower costs and emissions, higher reliability, greater self-sufficiency, and employment within the community. Since they run synchronized with the utility grid and their generators are small compared to the utility, failure and maintenance of any single microgrid generator is usually unnoticed by customers.
This article’s author, Ted Borer, is the energy plant manager at Princeton University and co-chair of the Microgrid Resources Coalition.