Is IDROP the forgotten engine for next-generation demand response?

,

If you set out to build a car, you wouldn't forget the engine.

Yet many utilities are thinking about virtual power plants (VPPs) without any idea where to get the engine to run them.

SGN's Discovery Showcase is devoted to smart grid technologies that are new or under appreciated. This installment concerns software in the latter category.Integral Analytics’ IDROP software has been around for years. I'm bringing it to your attention now because the electric power industry is going to require powerful analytical engines as we enter the era of Demand Response 2.0... but many of us seem to be forgetting this important requirement.

The challenges of demand response at scale

Our industry has been making some big claims for the value of DR. This will be even more true as we move away from simple load shaving and shifting towards more complex (but more valuable) load shaping.

As renewables become pervasive, as the wind surges up and down at our wind farms, as clouds drift over our solar arrays, we’re going to need a sophisticated engine to stay a step ahead. And to prevent those intermittent renewables from bringing the system down. Either that, or we will have to resort to lots of natural gas peaking plants standing by just in case.

The magnitude of the problem

You get a sense of the scale of the problem when you realize how much data the system has to juggle. First, think about all the data from smart meters and smart sensors.

Then think about all the data that needs to be taken into account for the network -- historical usage patterns, weather forecasts, wind and solar forecasts, transformer characteristics, feeder characteristics, distributed storage characteristics and much more.

Then think about the data for each and every customer:

• Customer parameters (such as historical usage and likelihood of opting out of a DR event)
• Comfort parameters (for instance, how much temperature variation each customer will tolerate)
• Appliance parameters and duty cycles (for instance, how often you can cycle an air conditioner without voiding the warranty)
• Electric vehicle parameters (time to a full charge, time until next likely use, owner preferences, etc.)

Imagine that mountain of data. Now imagine analyzing and optimizing it in real time. It’s the ultimate balancing act. It will eventually require us to fine tune millions of appliances minute by minute while also responding to hundreds of thousands of fluctuating solar panels and wind generators.

One approach to a solution

If I had to describe IDROP's goal in a phrase, I might say managing demand like supply. Today, utilities manage surges in demand by dispatching more supply. IDROP was designed to proactively "micro dispatch" hundreds of thousands (someday millions) of individual distributed  resources, including those on the demand side:

·         Individual appliances

·         Electric vehicle chargers

·         Distributed solar

·         Distributed storage

.

IDROP claims to be the balancing engine that can take everything into consideration and then adjust everything on a moment-by-moment basis. If I had to describe IDROP’s claimed technical differentiators, they would include:

·         Near real time (five-minute intervals)

·         Advanced optimization algorithms

·         Considers more parameters when making its decisions (including the actual cost of serving each customer)

·         Optimization down to the individual customer level

There’s another feature that will be of special interest to utilities still under a "traditional" business model. Under that model, demand response reduces the number of electrons sold, thereby reducing the utility’s revenues. Integral Analytics claims IDROP can reduce the peaks while keeping the overall load about the same. The result: The utility sells as many electrons as before, while reducing the number of expensive peak electrons it needs to supply.

As if that’s not enough, IDROP’s makers claim it can also handle storage dispatch, smart charging for electric vehicles (EVs) and bill targeting (the customer sets a target and the system adjusts comfort settings to achieve that goal).

.

But can it scale?

That’s all fine, but of course the trick is to do this at scale. If you believe (as I do) we’re moving into the machine-to-machine (M2M) world, then you realize that every home and office will contain dozens -- if not hundreds -- of individual devices all intelligent and interconnected. If you believe in renewable energy, then you realize every utility will be coping with hundreds of thousands of rooftop solar panels, not to mention intermittent wind farms. And if you believe in electric transportation, then you realize they will also be juggling tens or hundreds of thousands of EVs.

That's an astonishing amount of data to crunch and correlate. Is IDROP up to the task? The company says yes, but the jury is still out. Integral Analytics has participated in three pilots since 2008, most notably a small virtual power plant pilot at Duke.

Others are working on the same problem. (In fact, I will be back soon to tell you about an interesting new contestant from a surprising source.) Use the Talkback form at the bottom to nominate other virtual power plant engines that deserve to be discovered.

You might also be interested in …

Smart grid technologies: more on demand response

Demand side management news and trends

Stay connected with SGN …

Drop by our Smart Grid News Talk forums

Get LinkedIn with Jesse

Join Smart Grid News on Facebook

Follow us on Twitter

Try our RSS feed

Get our email digest