Storage and the Failure of Energy Markets.

As we build markets between intelligent buildings and the intelligent grid, there are two paths we can take to demand-response. One imposes central control, and is essentially the voluntary acceptance of a specific brown out with some local control of timing. This mode is traditional, as when the power company turns off the home hot water heater. The other path creates market mechanism to encourage local energy arbitrage between time-dependent markets. I’m betting that the second offers considerably more potential.

In today’s public concept and in the political world, the simple load shedding model is the only one. When you get a price signal, you should turn off your A/C. No hot showers in the afternoon. We will be happy to sweat in the dark because we will be saving the planet. Belief in the sustainability of this model right up there with the sustainability of the economic model posed by the red-suited man at the local grocery store this last weekend. It is a desire for how one wishes things worked trumping how things actually do. Very few people not already in monasteries will commit to long term privation, however slight.

Arbitrage is a little harder to think about but easier to sustain. Arbitrage between time markets lets buildings not only respond to times of congestion and shortage, but to also take advantage of times of abundance, to use resources that currently go to waste.

In simplest terms, arbitrage between time markets is the decision by the building operator to buy power when it is cheap, and to refrain from buying power when it is expensive. It no longer matters how inefficient local storage of energy is, just whether that inefficiency is less than the price differential. If I can make 50% return on investment by buying at night rather than during the day, then a 40% loss due to storage inefficiency is inconsequential.

Early adopters will pay a premium for this ability, but will also earn a premium as one of the few able to shed the load. As the market develops, both premiums will come down, as storage becomes more efficient ad as the market broadens, driving prices at both times toward the mean.

The largest barrier to large-scale adoption is the systematic under-pricing of power during peak use. This failure of pricing is yet another symptom of a market that has always foisted it costs off on others. Dirty coal plants push the costs of generation off on those downwind. Storm recovery costs push poor capital allocation and sub-market labor contracts into cost recovery allowed by the utilities commission. Brownouts push the costs of poor pricing strategies onto the general public; except during system failure, every brown-out is a failure of pricing. These systematic failures are predictable symptoms of an industry that has been a regulated monopoly managed for cost recovery for the last 80 years.

Proper pricing will validate and reward the energy storage market. Owners will embrace energy storage when they can see clean numbers from the market. Energy storage will flood the market when the owners want it.

Energy storage is coming. I ask you, my readers, what energy storage strategies do you think will be the early winners? I have some ideas that I will share this weekend.

The Case for oBIX in Laboratory systems

Well, if not oBIX, something like it.

Most data in modern research is collected by automated systems. Computers assess, quantify, and print out data. Some may be able to produce spreadsheets. Some produce graphs. (I remember measuring graphs with great care to turn them back into numbers in a previous career). Some may extrude CSV (comma separated values) files, to be imported into databases. But almost everything starts with machine measure and tabulation.

Often the information you need to understand the experiment has been recorded, but it is only available in a nearby system that the researcher either has no access to, or does not know he can get.

The biologist who recently asked for access to our operations data is one example. He works with plants in greenhouses. Greenhouses strive for, but do not always produce, specialized conditions. Understanding plants, and small differences in growing plants, often involves understanding the conditions they grew in precisely. We have a system that tracks minute by minute the temperature and humidity of the areas it monitors. In areas in which natural lighting is being used and augmented, light levels are also tracked. This researcher asked for access to the minute by minute temperature, humidity, and light for each zone in the greenhouses. What is wonderful about this request is that we can provide it essentially for free

There other systems, specialized systems that researchers work around. Back when I worked in a Biochemistry lab, we had large variances in the reactivity of the materials we worked in. After half a year, I guessed that these problems were caused by variances in the level of liquid hydrogen in the large carboys we stored samples in. Today, those carboys are replaced by ultralow freezers. While I never could prove this, I got outstanding results by working straight through (a 50 hour shift every other week) and thus eliminating variability. You can find the results on the web if you are interested in quassinoids.

There is a large cancer research center on campus. As part of the background for each grant that it submits, it includes general material on the quality of the facilities and how they enhance the research performed therein. One of the pillars of quality is an ultralow freezer tracking system.

This system monitors all the laboratory freezers in the building. Data on the quality of the freezer systems is carefully monitored. Each freezer has its tolerance, and it can be documented that each system stays within its tolerance. This documentation is part of the overall facilities quality report. If your sample is stored here, it will not be accidentally thawed out.

What is not available is any easy way for researchers to access the same information. If they ask the right person, they can get a spreadsheet describing the details of the performance of a particular freezer. They can request these periodically. This system, like most building monitoring systems, has no way to let researchers pull direct feeds of data from the freezer monitoring system. There is no way for one of several researchers using the same freezer to set personal alarms on conditions that matter to him. Most of the researchers are unaware that they can ask for anything.

Every automated laboratory system produces its own special data format. There is an effort underway, UnitsML, that hopes to establish a data standard for every measurable physical condition. All testing runs, all data, will be able to be delivered in a standard UnitsML format.

When the researcher can freely get to information on conditions whether from Laboratory equipment, from specialized laboratory infrastructure, and from buildings in self describing formats over the internet, then better analysis is possible. When students can use this information as well, without too many people interfering with experimental conditions, than education is improved.

I could go on…but the motto of out Building System integrator seems appropriate here. “No Data Left Behind!” These words are good for research as well as for building operations.

Biological Patterns for Systems Control

Last weekend, Fred Hapgood blogged at CIO magazine about network management and monitoring. He described how current models are hitting a wall of complexity and numbers. Conventional networks, and particularly their management, do not scale gracefully. As networks get larger and more complex their management problems will keep get more difficult even faster and the time scales for solving problems get smaller.

Hapgood points out in his blog that many researchers are now looking toward biological models for management and control. Biology is rich with large networks—protein cascades, gene switching networks, intercellular networks, nervous systems, and whole ecosystems that efficiently organize a large number of unreliable and dynamically changing components. These networks manifest adaptive and robust behaviors, despite the lack of any central management. This robustness and tolerance of diversity is in sharp contrast with man-made networks despite embracing far more individual variation among its nodes.

Hapgood went on to cite recent work primarily from the University of Bologna that tried to develop taxonomy of the modes of biological signaling and how they might apply to intra-node network communications and control. These modes of communications handle increase and loss of nodes well; more important, they degrade well, providing reduced functionality rather than failure. Fred was kind enough to share the paper with me (and I have placed it here).

The paper classifies biological signaling patterns into plain diffusion, reaction-diffusion, proliferation, and stigmergy. It goes on to consider biological entities as instances of object oriented design; and the signals as design patterns. Through simulations and modeling, they demonstrate effective and performant control of large systems. I would have liked the authors to reach for one more abstraction, to consider invoking these patterns in aspect oriented design. It is a fascinating and useful article.

Austrian school economics and developmental biology have long swapped concepts and vocabulary to describe the development and behavior of as complex adaptive systems. I think we are, as Fred suggests, beginning to recognize networked engineered systems as complex adaptive systems with the capability for their own emergent behaviors.

Complex adaptive systems have large numbers of diverse agents that interact. Each agent reacts to the actions of the other agents and to changes in environment. Agents are autonomous, using distributed control and decentralized decision making. Eventually, the dominant interaction becomes the agents interacting with the system environment that was itself created by the agents’ own independent decision making.

In economics, we call the order that arises out of markets emergent self organization. In biology, we call it embryology. In either case, a large scale pattern emerges out of the smaller decisions and interactions. The emergent pattern is not imposed top-down, but rather arises from decentralized agents interacting within bounds of distributed control (or self control if you will).

A characteristic of meta-systems (or systems of systems) that demonstrate self organization is resilience in the face of change, what the economists call adaptive capacity. Market design theory, in the news this week with a new Nobel Prize, is in part concerned with ensuring adaptive capacity.

We are just beginning to apply the concepts of biology and markets to aggregates of engineered systems. In nature, systems that have too many direct interactions become brittle, and break badly. The Cleveland Outage of 2001 could be described as such a shattering, with the cracks extending into Canada and the East Coast. Less control and more heterogeneity in agents may be what we need to acquire resilience in our engineered systems.

How do you buy a Green Volt?

Several readers have written to me contesting the entire idea of Green Power, and buying from the producer of choice. Electrons are Electrons they say. They all come down the same wire. It makes no sense to try to buy green power.

I assume these people are also not participants in the modern economy, and barter for all goods. Money is a way to transfer value between multiple producers of commodities more efficiently than direct trades. I don’t need to accept direct value in turnips for my work. The banking system takes this a step further. I get my dollars and put them in the bank. This allows me to write a check and send it to you. You take that check to some entirely different bank and get cash. Neither of us worries whether your dollars have the same serial numbers that I deposited in the bank.

Buying green power should work in the same way. Green producer puts volts onto the grid. I buy volts from the grid. Electronic transactions mediate these processes. Neither of us needs to worry about the quantum serial numbers on each electron.

There is an old parable about Stone Soup. An itinerate comes to town, and asks not for food, but for a pot and water to make stone soup. He places a stone in the water and begins cooking. Soon he tastes it and proclaims that it is wonderful soup, but would be better if it had some carrots. Overcome by curiosity, one of the townsfolk produces some carrots. In the same way, he soon gets some potatoes, and cabbage, and onions, and so on. All the time he extols the virtues of stone soup.

The problem is, today, the doubters are closer to correct. My power company has a portfolio that they claim is green power. They augment it with power from other sources because there is no way to temporally allocate that green power. I, as the customer, have no way of knowing if they have sold the output of a single windmill several hundred times, or if the price is sufficient to encourage adding more windmills. I am suspicious of paying a premium for green power—it may be just stone soup.

If there were markets, then I could buy green power at a higher price. There are those who would do this just as there are those who buy expensive potatoes at Whole Foods. Because of Whole Foods, there are now many more producers of organic potatoes—and no one suggests that the industrial potato farms should get a share. In the same way, honest markets would incentivize green power far better than do today’s regulated offerings.

Some people see no sense in paying higher prices for such power. Some may even argue that flying organic raspberries in by jet from Chile in February may not actually save the planet. It doesn’t matter. In that market, people have choices, and can buy in accord with their values. These purchasing decisions would be reflected directly in Alternate Power profitability (and thus in Alternative Power investment).

With green power comes some reliability issues. Perhaps I want to purchase reliability assurances from a generator consortium. Perhaps I want to store power locally to tide me over. Either way, markets will offer me solutions if we let them.

We have the technology, today, to make each choice available. We have the technology to make these choices transactive, including support for time of day or dynamic congestion pricing. Installing such technology will be cheaper by an order of magnitude or two than building out the transmission infrastructure of today under the cost recovery innovation-averse regulations of today.

And it will create a market wherein I can choose the green power I want, just as I can choose organic raspberries in February. I won’t need to justify it to a utilities commissioner or to an engineer. I’ll be able to do it because I want to. And it won’t be because I like stone soup.