Demand Response Management in Terms of a Blowfish, Batteries and Paint

I like a good visual aid that conveys meaning and distills complex data into an elegant illustration. In the continuous spectrum of utility demand response management (DRM), so far no one has modeled the problem in terms of a blowfish, some batteries, and a little paint. What if such a model could help us understand how a utility could, in effect, sell capacity to itself, and help leverage renewables more efficiently? Indeed, this would be a rosy picture.

In his landmark treatise The Visual Display of Quantitative Information, Edward Tufte demonstrates how otherwise voluminous, mind-numbing information can be rendered into an illuminating image. An effective interpretation helps us become more intelligent, and saves considerable time and frustration. My model starts with something most of us already have a good mental picture of: a temperamental fish you wouldn’t want to grab.

Demand response used to be modeled as binary game: either 1) overproduce capacity or 2) reduce the inevitable and often roughly predicable daily peak demand. Satisfying demand was sometimes a matter of overkill, like driving a finishing nail with a sledgehammer. In part due to the exponential growth in renewable energy capacity, energy storage adds a third, stabilizing wheel — if it’s approached intelligently.

I’ve been subjected to many DRM diagrams, the vast majority of which are a variation of a sine wave. Some depict data such as cost, power mix, emissions, peaks, troughs, etc. If anyone absolutely must draw me another whiteboard DRM diagram with similar shape, they should at least have the decency to get an aromatic dry erase marker to hold my olfactory interest. I like the cherry-scented one.

Westerners may know a blowfish (or fugu in Japan) as “that fish that can poison you to death with a sloppy fillet job.” Just as a clumsy sushi chef can ruin your day, bad DRM by a utility can ruin a lot of peoples’ day.  However, the topography of a temperamental blowfish is not only a great analogy for the load, but it also a model for the most intelligent response.  Understanding contours gives depth to any map.

Saving for a Rainy Day

The crucial third wheel of managing demand is the optimal use of storage technology. Three wheels make a tricycle more stable than a bike. Popular renewable sources, including solar PV, solar thermal, and wind (arguably just another form of solar energy) are inherently intermittent, a fancy way of saying unreliable.

Utilities are exploring and deploying methods of harnessing unpredictable nature of some renewable energies and transform them into constant, predictable sources of capacity. Some storage technologies involve saving energy in large electric batteries, industrial grade flywheels or by geologic or hydraulic means.

I’ll submit that large-scale storage is an incomplete solution for a 21st century grid, especially as renewables constitute an ever-increasing portion of energy portfolio. To properly greet the new energy millennia, we’ll have to start dealing with storage on a microgrid level. Comprised of many small-scale energy sources, semi-autonomous microgrids have many attractive properties to utilities, regulators, investors and consumers.  Microgrids are the new black.

 Modeling Demand with a Blowfish

Demand cannot be viewed in one dimension, any more than speedometer can tell you what direction you’re heading. Past a certain point, demand observation and prediction are more magic than science, which is one reason why utilities trade energy, often at unfavorable cost.

A blowfish is remarkable in that it’s rather cute when in a calm state. When excited, more specifically when threatened, it puffs up and its tines are ready to poke. It’s not a complex beast, even if it has a fragile temperament.  Also, it possesses striking contours that can’t be described by a simple mathematical formula.

As our starting elegant image, let’s pretend that a blowfish is suspended in the middle of a sufficiently inflated latex party balloon. In this model the fish represents the demand, and the balloon the capacity. If we wish to satisfy our pointy chum and keep our system in equilibrium, we’ll have to keep our outer capacity balloon inflated to prevent the threat of puncture. 

Even at calm moments when the fish is small, there is significant inefficiency.  Waste can be expressed by the gap between the balloon and the fish, specifically by the gross volume of the balloon minus the volume of our somewhat unpredictable shape-shifting pal. Inefficiencies such as wasted dollars, unused electrons and extra carbon are quantitative.

When our blowfish gets agitated and increases in volume, the demand must respond in kind. The balloon must get larger, proportional in both size and rate of expansion, which can be a tall order. Again, overcapacity is the traditional utility reaction. If we stick with our volumetric model, the larger the fish, the greater the potential for waste. The risk of underproduction can be qualitative, like light switches that don’t work, customers bristling because they can’t watch American Idol, and grumpy PUCs.

It’s easy to see that the better our knowledge of both the fish’s changing size and contours, the better our ability to appropriately meet demand. It would be great to react to demand at the speed our uppity fish.  A blowfish changes its shape rapidly because it doesn’t want to become an hors d’oeuvre for a predator between lunch and dinner.  Likewise, consumers don’t want to be victims between lunch and dinner, when peak demand usually occurs.

Spray Paint and Duct Tape

I once had a boss that claimed that any mechanical device could be repaired by paint and duct tape. Needless to say, he wasn’t a very good boss. But, what if we slap some paint on our blowfish? Or, more exactly just shy of covering our fish in paint, we approach the limit of paint, or virtual paint on the beast — paint that floats just above this unusual shape? The volume encased by this paint and the volume of the fish that is underneath are nearly identical, with a little breathing room.

Imagine that this paint is infinitely flexible. The approximated paint wetsuit allows us to navigate every contour on our fish. And by mapping the contour to the most granular level, we will better come to understand how his demands change.

The blowfish does not have to solely represent the demand of a utility’s entire domain. It can also represent subcomponents, or microgrids, each consistent of multiple smaller entities that represent a subset of physical (geographic) or logical (such as consumer type) demand. Even if the utility’s visibility into one residence is not to the inside, or home area network (HAN) level, there is still useful real-time demand information to be aggregated from statistically significant segment of consumers.

To a point, the smaller the microgrid, the easier it might be to manage shifting demand. If only we could respond with capacity as fast as the speed of light.

A Battery S.W.A.T. Team

While responding at the speed of light is unfeasible, responding at the level of information is a more likely proposition.

There are several reasons why battery technologies make sense on various scales, but there are many considerations in the size, type, location and features of any stored capacity that would be used on a microgrid. If local demand data is in near real-time, then it follows that the response to this information should be as rapid as possible.

One example of a large-scale transmission grid storage device is the sodium sulfur (NaS) battery.  Like any battery technology, it’s evaluated on a host of criteria including energy density, size, cost/watt, nominal voltage, total lifecycle, environmental impact and many others.  NaS cells have been tested in an automotive setting, and seen some action on the space shuttle. The problem is these batteries are large and hot, which does not make them suitable for every application or every setting. For NaS, substations are a fine locale, neighborhoods not so much.

A much more promising technology is rapid charge/discharge batteries, for several reasons.  First, assuming they meet certain industrial-level criteria, they can be deployed as part of an array of capacity nodes in the distribution network.  Their potential to help rapidly micro-manage localized demand makes them amazingly attractive. Also, such an array would have advantages in redundancy and outage management. And finally, they can take greater advantage renewable generation during both solar and wind peaks, and minimize the grid loss at peak demand inherent in centralized capacity storage.

One promising technology is a new generation of lithium iron phosphate (LiFePO4) batteries (also known as LFPs) being developed by Gerbrand Ceder, a professor of materials science at MIT.  His prototypes, made of a compound already in use in early versions of GM’s Volt automobile, charge at rate an order of magnitude more quickly than today’s production materials of the similar composition.

Next generation LFPs have other favorable attributes, including charge retention, energy density and non-toxicity. While they will likely not follow Moore’s Law of form-factor reduction over time (according to Ceder), these batteries still have some opportunities for greater energy density before they would be deployed in a few years (for more, listen to NPR’s Science Friday from March 13, 2009). If this new generation of LFPs (or any other newcomer) can achieve favorable criteria for the utility scale capacities, this could bode well for adoption within microgrids.

Rapidly shifting capacity from one microgrid to another maximizes the ability of the network to adjust.  Adjusting at off-peak hours to leverage renewables is one advantage, but also smart sharing during any condition can make the most of capacity, peak or not, renewable or not.

There’s another advantage to rapid charge-discharge microgrids. Assuming a utility has sufficient battery-stored capacity on the transmission grid during a demand spike, there is no guarantee that capacity can be dispatched, or used effectively.  In the case of a utility capacity trade, the laws of physics may interfere with delivery.  Within a utility’s own domain, line loss takes its usual cut.  Delivery is least expensive when two criteria are met: 1) capacity is obtained when demand is lowest (least cost) and 2) capacity is delivered in closest proximity to demand (least loss).

The Celestial Spheres

Early in the 16th century, astronomer Nicolaus Copernicus disturbed the status quo by removing the earth as the center of the universe. Until then, the battle cry of the masses was much like that of the unimaginative office worker: “but we’ve always done it this way!” There are few expressions I loathe more.

In the most traditional sense, the roles and expectations of utilities and consumers have always been a certain way as well. The ideal utility was somewhat like an NFL offensive lineman, largely invisible until a penalty flag is thrown. Consumers were typically satisfied if a light would come on when the switch was flipped. We no longer have the luxury of such nostalgic thinking.

Renewable energy technologies radically change our notion of capacity, and utilities by and large have been lethargic in response to them.  While 50 experts will give you 100 answers on what the Omniscient Grid will be, none could disagree that our infrastructure is antiquated.

Former Sun Microsystems Chief Researcher John Gage is credited with the phrase “the network is the computer.” This represented much of what was to come in the computing world, basically the concept of a virtual, on-demand supercomputer. Likewise for the grid, the power is in how responsibility is shared amongst the components. Shuttling power rapidly on a microgrid for peak response (or outage management) parallels existing functions on the internet, giving us hope that the sum of the grid can likewise be greater than its parts.

Sharing power between member microgrids can also be thought of as a form of bartering. Rather than utilities trying to figure out how to sell power to each other, they would be figuring out efficient ways to sell it to themselves. Ultimately, this serves the customer better.

An efficient, lithe and agile utility needs a DMR program that relies on granular information, redundancy and effective routing of capacity.  If there is an elegant way ensure demand is met without compromising revenue, then utilities will be attracted it, especially if it’s in their control.

Back in the dawn of the space age, some futurists predicted that we’d be having picnics on the moon by 1977. I would have liked that, but it would have been a luxury.

Until Copernicus piped up, the common belief that the earth was surrounded by concentric celestial spheres (such as the moon, the sun and the planets) was not widely challenged. It was just the natural order until reality got in the way. A 16th century picnic on the moon would have given us a nice picture of how big things and small things can play together a little better.

Jim Turosak is an independent energy researcher, analyst, software engineer and ultramarathoner in Denver, CO.

Previous articleSolar shootout in the San Joachin Valley
Next articleSiemens In US Wind Turbine Production Plan

No posts to display