America's power grid has changed a lot recently. For more than 100 years, we relied on large centralized coal and hydro, and later nuclear generating facilities, to produce our electricity. Now, as we move into the 21st century, centralized electrical generation is being gradually retired, forced out to make way for the new kids on the block - renewable and distributed energy resources (DERs), like wind and solar.

What's evident is that the move to DER is irresistible and unstoppable. After all, why would anyone not want to use the wind and sun as primary generation resources if it is economic to do so? The fuel is free, in-exhaustible, non-polluting, and creates no carbon emissions or dangerous nuclear waste.

Indeed, people do want this. Several US states have set ambitious renewables goals. Hawaii, California, Washington D.C. and New Mexico have all enacted legislation to eliminate carbon emissions from the grid and move to 100% renewable electrical generation.

The grid operates in such a way that small levels of DER penetration can easily be accommodated, but as DERs become a more significant part of the generation portfolio, problems start to arise.

These problems are mainly to do with the fact that the grid has to operate in a steady, instantaneous balance of supply-and-demand. This could be achieved when generation was highly predictable and could be called on to increase or decrease in lock-step with the total demand. Not so with the wind and the sun. And whereas a coal or nuclear plant can happily run round-the-clock, the sun can only produce meaningful levels of photovoltaic power for about six hours a day, and the wind only when it's blowing.

The old control systems that kept the grid in balance were designed to operate under the old paradigm --- highly controllable, always-on generation, and a total load that was highly predictable over the course of any given day, with some slight variation caused by how relatively hot or cold it was compared to the same day yesterday.

These control systems were not designed to operate in the intermittent world of renewables, especially large amounts of DER located out at the edge of the grid, and this issue presents a major barrier to the ambitions of these states. If DERs are to ever become a truly significant part of the generation mix in America, a new type of control system is needed. These new systems are referred to as Distributed Energy Resource Management Systems -- or "DERMS".

So, what is different about these DERMS, and how can they help DERs play a bigger role in America's energy system? To answer this, we need to look at the penetration of DERs on two levels.

First, let's look at large-scale DER deployments, the equivalent of a large nuclear or coal plant. Think of wind farms with hundreds of turbines on the central plains, or the huge desert solar projects deployed in California. These are often referred to as 'grid scale' renewables, and generate many hundreds or even thousands of megawatts. They are typically connected at the high-voltage transmission or sub-transmission level.

To the grid control systems, these farms or plants look like other baseload generation, except that like hydro plants they are "resource constrained" -- the sun sets in the same way as a reservoir feeding a hydro plant might start to run low on water. They can be intermittent: less reliable than traditional sources.

What the DERMS really needs in order to manage these resources is an accurate forecast of the wind or solar power available over the day. Traditional grid control systems needed an accurate forecast of the demand, or load, while the generation available was (usually) a given. In the presence of high penetrations of intermittent DERs, the DERMS systems have to have an accurate forecast of the amount of solar and wind generation that will be available, in addition to the load -- a new requirement.

Much work is being done to try to enable accurate forecasts of this kind. At least we know with a high degree of certainty what time the sun rises and sets, but the amount of cloud cover at any time? That's a more difficult problem.

Next there are the smaller, but still significant, megawatt-scale DER deployments. For example, a farmer who has converted an unused field in New England into a solar farm, or the Walmart in Kansas that has covered its parking lot in solar panels. In addition to these, there is the increasingly common DER deployment at the consumer household level -- around 5-20 KW each. Both of these types of DER can (and often are) connected to the grid at lower distribution voltages.

The new "solar-plus-storage" mantra is everywhere and for good reason. People now have the chance to power not only their house but also their commute vehicle with the free, non-polluting fuel provided by their own rooftop solar system.
b These mid-size and consumer-size deployments present new challenges to traditional grid control systems. They are often connected beyond the "visibility" of the old, centralized control systems, and generation can swing very quickly, creating local grid stability problems. The predictive modeling used for the bulk power grid, which works well in the presence of good telemetry of power measurements and huge levels of system inertia, is ineffective when large amounts of DERs are connected at the grid edge.

There is also little opportunity for operator control, as events at the grid-edge can occur too fast for an operator to intervene effectively, and anyway there are simply far too many locations for an operator to monitor effectively. This is where the new DERMS systems come in. Large deployments of DERs, at the level required to reach 100% carbon-free renewable generation as demanded by more and more states, require automation at the grid-edge in the form of distributed computing systems that can act on their own -- autonomously.

These distributed systems have to be locally aware -- what is connected, what the stability problems might be, what action can be taken for over and under generation situations, feeder-by-feeder, substation-by-substation. Batteries are a tremendous resource in these situations, being able to switch almost instantaneously from consumers to generators. But the operation of the batteries as the rest of the load generation fluctuates locally has to be tightly coordinated, and the control systems have to be able to do this on their own with no help from an operator.

The old paradigm of centralized systems with little visibility of the grid edge (because that was mostly just predictable consumption) doesn't work anymore. Massive quantities of grid-edge DERs are being deployed and we need new ways to control it. This is not to say that centralized systems can't still be useful. DERMS systems are typically connected to and communicate with these centralized systems, which can then perform wider-region optimization of overall grid control. Act locally, think globally applies.

Millions of DER located at the grid edge create problems for traditional utilities as we have discussed. However, we should also consider their value to the utility. Lots of small DER can be grouped together to create a distributed, resilient and bi-directional (generation and load) Virtual Power Plant (VPP) of significant aggregated scale that the utility can call on when needed for multiple uses. Here the intermittency of individual DER is "averaged out" by the large number and their distributed geographical locations.

For instance, traditional demand response programs required participating sites to cut power for a while, but they can now switch to clean, local generation, or increase battery charging when the utility has excess power on the grid. For the VPP "owner", there are potential revenue streams here, and also through participation in the energy trading and ancillary service markets operated by the ISOs. Units traded here are required to be a minimum size, ruling out DER from small participating sites, but not for an aggregated VPP. However, this requires systems and solutions that can aggregate individual DER and represent them as a single tradable entity, or VPP. Today, most VPP solutions focus on larger industrial and commercial customers, but this is changing as these DERMS systems start to be deployed.

Indeed, DERMS systems capabilities extend beyond the traditional VPP, by also enabling fast acting, critical services that require back-ups and fail safes, and locational network services that utilize a diverse set of DER blended together on a single feeder. And there are new systems out there that enable significant quantities of DER to be added to these feeders -- far beyond what traditional utility models would normally allow -- due to their ability to react quickly and autonomously to local events in real time.

Ultimately for the U.S., change is coming. We are seeing it happen already. In June this year, New York became the latest state to announce ambitious plans to operate 100% clean energy by 2030. Fortunately, DERMS can help utilities prepare for this change while paving the way for DERs to play a pivotal role in America's future energy mix.

About the Author:

Peter (Pete) Maltbaek
Executive Vice President and General Manager-North America
Smarter Grid Solutions (SGS)

Peter is a member of the Smarter Grid Solutions (SGS) global executive team, responsible for the management and growth of the company's business throughout North America. SGS has spent ten years developing, deploying and proving a unique approach to managing the Smart Grid, and is recognized as a thought leader in this domain. SGS has worked with and learned from electricity distribution companies, regulatory authorities, university research teams, DER developers, SCADA/DMS suppliers, grid edge device manufacturers and many others. Prior to his current position, Peter held senior-level management positions at Nexant, Power Assure, Silver Spring Networks, and CPower.