Connecting generators to the grid is technically involved and expensive, so much so that the price tag and impact assessment can be deal-killers. Resulting interconnection charges, notes Robert Lasseter, professor emeritus of electrical engineering at the University of Wisconsin, "can sometimes run as much as the distributed generation [DG]" itself, and this is obviously prohibitive. To be allowed a connection, a generator must also comply with the Institute of Electrical and Electronics Engineers' (IEEE) standard P1547, which requires automatic and rapid disconnection in the event of the grid experiencing a fault. Naturally, this is a great inconvenience to the DG owner.

Lasseter is part of a major initiative that's now working hard to solve and perhaps forever alter—dramatically—the connective relationship between DG and utilities. Throughout 2004 his Wisconsin laboratory was the site for proving the needed breakthrough technology. Lasseter is participating in a group called the Consortium for Electric Reliability Technology Solutions (CERTS), a major R&D initiative formed six years ago to find ways to improve the nation's electric power systems. CERTS members include Lawrence Berkeley National Laboratory, Oak Ridge National Laboratory, Pacific Northwest National Laboratory, the National Science Foundation's Power Systems Engineering Research Center, and Sandia National Laboratories. Their work is currently being funded by grants from the US Department of Energy (DOE) and the California Energy Commission's (CEC) Public Interest Energy Research (PIER) Program.

One primary focus—potentially the most revolutionary for DG—is the group's emerging technology and design concept for electrical microgrids. First, by way of definition, a microgrid consists of two or more DG resources on a feeder connectable to a main grid; this allows it to run either in parallel or independently of the grid. Similar configurations in various flavors have existed on campuses and military bases, etc., for some time, of course, but CERTS's innovation promises to extend microgrid versatility dramatically. CERTS has published several white papers about its model, from which some of the following report is adapted.

Enjoying Common Connections
Lassiter summarizes the CERTS concept. Instead of having multiple, separate DG-to-grid connections, each governed by IEEE P1547, all DG resources are tied together on their own new feeder, which is then linked to the grid at a single point of common coupling (PCC). The main grid is no longer directly exposed to the individual DG voltages and outputs; instead, the feeder, with its "microscources" of DG power—now comprising the microgrid—presents itself to the main grid "as a single, very well-behaved dispatchable load," says Lasseter. A microgrid might thus incorporate several or even dozens of DG resources and many loads—and yet, he reiterates, "instead of each having to be independently grid-connected, they are all part of the microgrid, which has a single connection."

Moreover, he adds, at this PCC, newly emerging automatic switching technology will mean, "We have the ability to intentionally island [disconnect] and reconnect." Here, the chief advantage for DG is a kind of permissible circumvention of IEEE P1547 and its automatic shutdown requirement. Instead of forcing DG to cease making power whenever the grid's voltage begins wavering, all the microgridded DG can keep running, because they're on their own detachable feeder. That feeder or microgrid islands at the PCC, thereby protecting the main grid from any errant voltage coming from the DG. The latter may continue running without interruption—obviously, a great convenience for the owner. After the grid problems ease, the microgrid can then seamlessly reconnect.

Benefits of CERTS's innovative approach actually accrue to both the main grid operator and the microgrid community (see Benefits for Power Customers on page 24).

Given the numerous operational advantages of CERTS microgrids, the IEEE is also now writing a new standard—dubbed P1547.4 and expected to be issued sometime in 2005— which will propagate appropriately updated rules on how microgrids and other resources may island.

Surmounting Several Challenges
How exactly will DGs on a microgrid be strung together? In order to answer this, it's first useful to be aware of three technical challenges posed and how the CERTS design solves them.

How to Disconnect and Reconnect Seamlessly
Disconnection isn't a new problem for DG, of course, but because a faultless and smooth transition is so critical to the microgrid concept, certain switch-system modifications were deemed necessary. The Wisconsin lab developed a proprietary static switch device employing "back-to-back SCRs [silicon-controlled rectifiers or thyristers] with local logic to re-synch them," Lassiter notes. The results meet EPRI quality power standards.

How to Compensate for the Big Power Loss During Islanding
When the microgrid islands, power from the main grid quickly detaches; this means DG must rev up and counteract that lost power. But DG isn't typically capable of doing this quickly enough. The solution? Stored DC power and even stored AC power. Each energy resource on the microgrid, whether AC- or DC-powered, outputs to a DC bus before being inverted (to 60 Hz grid-quality AC, at each voltage source) for energizing the microgrid. This power inversion rectifies the output and makes the diverse sources compatible. Positioned along the DC buses, assorted innovative storage devices can be applied, such as lead-acid batteries (now with greatly increased storage density and extended lifetimes); supercapacitors with very high discharge rates able to handle quick load shifts; and superconducting magnetic energy storage, which pass current along without losses and store energy within a superconducting electromagnetic coil. Advances in all of these methods enable quick energy bursts from storage. On the AC side, similar quickly dischargeable storage has been developed, primarily with extremely high-rpm flywheels. In any event, these technologies enable the microgrid to compensate with bursts of power as needed, within a few cycles, until the DG prime movers can power up fully—in about ten or twenty seconds. Meanwhile, rapid load-shifting is also likely to take place, again automatically.

How to Stabilize Voltage of Many Generators on One Feeder
Solving this turned out to be the most difficult part, Lasseter recalls. Voltages tend to sag variably, due to distances between machines and fluctuating currents, loads, and impedances. DGs then begin to "feed off against each other," he says, and try to compensate—causing instability.

To solve this problem, automatic voltage and power wattage regulators were added, enabling machines to adjust to shifting loads and keep the voltage steady. CERTS researchers had to build "droop" into the voltage controls, he recalls, "So they won't fight with each other or couple to each other."

Already built into a number of DG products are advanced power electronics. These include the inverters just mentioned, automated controls for load-following and power flow, and differential current circuit breakers. Some or all of these elements are incorporated, for instance, into Capstone microturbines and, reportedly, in the Honeywell 75-kW Parallon, as well as in generator products from Bowman and Turbec. Controls respond to local frequency and voltage readings, adjusting appropriately to maintain balance while load-following. Such a generator, explains Lasseter, "automatically ‘rethinks' itself without requiring communication" from a dispatcher. Capstones, in particular, use automated controls to operate in sync when they're arrayed with each other in series (i.e., forming a kind of "proto"-microgrid). Advanced electronics also support many photovoltaic (PV) products and fuel cells.

To enable these built-in electronics to do higher-level interactive functions on a future microgrid, the onboard controllers will need to be somewhat beefed up and modified. But the essentials are already in place for maintaining flow on the feeder, regulating voltage at each source, and even power compensating during islanding. Controls will also need to operate in two new distinct modes, grid-connected and islanded. Assorted control software and PLC modifications are already reportedly being prepared by one or two manufacturers. In any case, increasingly automated DG capable of self-adjusting in harmony with local resources and signals, will be critical to DG's microgridded future

During trial runs of a basic working simulation at the Wisconsin lab, smart PLCs and automatic onboard controls "proved out the system behavior" as Lassiter hoped. "That it's truly ‘plug and play' doesn't have to be tested," he reports: "It's working."

Real-World Demo Coming Soon
Quite soon it will all be working in an actual application. Groundbreaking for the first prototype microgrid, largely inspired by CERTS concepts, occurred in Waitsfield, VT, in late 2004. Completion is expected sometime in mid 2005, reports Jonathan Lynch, chief technology officer of the firm that is building it, Northern Power Co., also of Waitsfield. Northern Power, a subsidiary of Distributed Energy Systems Corp., has extensive experience in DG, having installed more than 800 power systems for industry and governments worldwide.

Northern Power has designed this first-of-its-kind microgrid to serve itself and its nearby corporate neighborhood, called Mad River Park, a light commercial and industrial center in a semi-rural setting. Electric power service is currently provided by the Washington Electric Cooperative (WEC) of East Montpelier. WEC is a collaborative partner with Northern and stands to benefit from the grid's mix of additional clean, renewable, and low-emissions generation. Northern Power also does R&D on renewable energy and hydrogen technology, and some of the activities on the new microgrid will be quasi-experimental; it's being supported by a DOE assisting grant of $600,000.

Northern's initial array of resources will consist of propane-fueled reciprocating engines, microturbines, a photovoltaic array, and a small wind turbine. Integrated with these are storage devices. The fuel-burning generators will cogenerate usable heat and power for maximal efficiency. And later on, Lynch notes, other important cutting-edge technologies will be tested, such as multi-fueled Stirling engines, fuel cells, and advanced storage batteries.

Cumulative power output will reach around 350 kilowatts initially—enough to support loads from five Mad River commercial and industrial facilities and 12 nearby homes currently served by WEC.

As for design specifics, says Lynch, the power sources will be separated on a single feeder at distances of at least 100 yards. Automated electronic controls on the inverters will do the "heavy lifting" of voltage and power balancing. The systems are pre-programmed with algorithms developed and tested at the Wisconsin lab. Because Mad Rivers' microgrid system is a prototype, it will apply additional voltage regulators for protection. Also, once it's operational, its fluctuations will be closely watched as loads and outputs are varied under real load conditions, and experimentally, to see what happens.

Power coming from the local WEC substation will also be monitored (automatically); a protective relay at the PCC will detect undesirable events. If grid conditions warrant, a fast disconnection switch will open. The microgrid will island and continue operating on its own, so that microgrid clients will enjoy uninterrupted power. Actually, notes Lynch, several modes of operation are possible with the sophisticated switch, including full islanding to insulate the neighborhood from voltage sags or spikes on the main grid, as well as variable kinds of parallel connection.

Future Commercial Applications
Northern Power hasn't announced formal marketing plans yet, but potential opportunities for these ultra-functional, highly beneficial microgrids are obvious. Lynch envisions them serving, for example, as attractive alternatives for new commercial/industrial parks where utility-line extension is often costly. Microgrids might also power remote villages in the developing world, "or in any other area where you don't have a strong utility, or you can't rely on it, or it doesn't exist," he says. Bypassing the need for a centralized, multi-megawatt generator, communities that are favorably sited in sunny, windy areas could easily tap these networks to create all or nearly all renewables-based power systems.

WEMPEC Microgrid Emulation Lab

Moreover, new kinds of business relationships can readily be foreseen. For example, a commercial power system operator might buy and run a microgrid-based DG network, but not necessarily be an end user itself. Rather, it would act as a "mini-utility" for a special niche. Another obvious scenario would be that of utilities adopting microgrid technology themselves, perhaps then partnering with cogen owners collaboratively.

Other intriguing but as-yet unanswered questions arise: How, for example, would peak-shaving economics be affected if the DG resources are microgridded and running all the time rather than optimized? And how might a cogeneration site share its power with its microgrid neighbors? Lynch points out that these and other technical, regulatory, and commercial issues remain virgin territory for markets and adopters to explore.

From a researcher's standpoint, Lasseter sums up by noting that more demonstration sites would be nice to have, especially in collaboration with utilities. As of late 2004, CERTS was already evaluating proposals from three such prospects, and has intentions of selecting and implementing one sometime in 2005. The CEC, he notes, will provide a generous grant of about $3 million to build a full-scale microgrid at an unnamed California site. Other potential partners for CERTS' energy solution might include business and industrial parks, campuses, military bases and office buildings-specifically, says Lasseter, those with loads up to half a megawatt at 480 volts (i.e., for testing a 500-kva transformer) and which are able to use waste heat. A prospective microgrid could likely integrate an existing energy management system, too, he adds.

Finally, part of the rollout process that's now under way involves demonstrating or at least probing the potential commercial value. CERTS, Lasseter says, "is talking with potential partners to take it to the next level."

La Mesa, CA–based writer DAVID ENGLE specializes in construction-related topics.

DE - March/April 2005