Cargo Cult Energy

I spent last week in Chicago and by of Silicon Valley, talking about new energy. In Chicago, we were talking about the smart grid, and how it enables new markets in energy. Out by San Francisco Bay, the conversation was, of course, about ventures and new businesses and high tech. There were exciting conversations in Chicago, ones that may lead getting the underlying structures of smart energy markets right. There were innovative projects in California, ones that are beginning to answer "What would your stuff do, if it knew the price of energy, now.?" In both locations, there was a tendency to fall into a trap that I call Cargo Cult Energy.

The phrase Cargo Cult names a reaction of some isolated islanders in the South Pacific to what they experienced in World War II. Some of these islanders had never seen internal combustion or manufactured goods or any food that they had not themselves pulled from the sea or hewn from the land. One day a stranger would come, or several. These strangers seemed very determined to cut down trees, and to flatten the land. The strangers were so obsessed that the islanders helped them, even going so far as to build a tower at the end of the flat space.

The strangers would go up into the towers and call down huge flying machines. All the supplies necessary for industrialized war would flow through this airstrip on an isolated island. The leavings dropped by the runways, and pilfered from the warehouses were more wealth than the islanders had ever imagined. The war ended, and the strange men left, and the flying machines came no more. On some islands, myths grew. If only the towers were maintained, if only the right rituals were performed at the end of the runway, then the machines, and then wealth would return.

In Chicago, fat too much of the conversation, before the GridEcon started each day, was of incentives. Over breakfast, alas, the conversation was often not of systems, and technology, and business process. Too often, plans were being built around short term incentives. What incentives do they have in New York? When do the tax incentives expire in Illinois?  We are not talking about priming the pump here. The business plans are short. Can we get in and get out when the incentives expire?

The venture capital guys were clear. They were not interested in funding any project whose business plan was based on tax credits, of utility rebates. What government gives, what the public utility commission grants, can just as easily be taken away tomorrow. Venture money wants long term value. Each technology should be sold on its clear and identifiable business value. Once that case was made, credits, and rebates could be a sweetener, a way to accelerate the business cycle.

Around the bay, I saw some many technology plans. I saw novel integrations of existing technology, in which simple things were made smart, particularly in how they used energy. I saw polymath projects, in which technologies and approaches from all over were combined into a novel product that used smart energy. There is a buzz of something ready to happen. Unfortunately I also saw folks tempted to lose their virtue.

Silicon Valley prides itself on a "virtuous culture of innovation", in which good products win, bad products lose, and hard work gets you ahead. I saw some very interesting, and perhaps some very good products. Too often, though, the management team forgets about the building the internet of things around energy, and gets lured by the siren song of third party programs. It’s great – they won’t even have to pay for it! We’ll pay for the installation with DR dollars! The homeowner won’t care because they’ll get a tax credit! We are not talking about priming the pump here. The business plans are short. "Can we get in and get out when the incentives expire?" In other words, these plans were without Silicon Valley virtue.

The energy markets in the US have been poor markets, looking to the regulators rather than to competitors for 100 years. To the extent that the smart grid enables new markets, successful new ventures will chase those new markets. Unless of course they get seduces by unnatural signals coming from the externa of the old markets. Unless they build their business plans around Cargo Cults.

What if the Smart Grid had Reliable Generation

Most of today’s conversations about smart energy have at their core recognition that new energy is inherently unreliable. That unreliability will flow throughout the grid, and those that rely on the grid (homes, buildings, industry, vehicles) will need to consider that unreliability. The requirements reach beyond the operation of transmission and distribution (T&D) to intelligent end-points, able to adjust energy requirements, store energy locally, and even supply energy back to the grid. This requires two-way symmetric communications.

But what if new technology provides us with rock solid reliable power? A couple reports this week have turned my thoughts to the problems of the smart grid given perfect reliability of generation

In the last week, I have read reports of working prototypes for traveling-wave reactors, able to rely on minimally processed fuel, and to consume their own waste, running for a couple hundred years without refueling. I have also read of small package nuclear power plants, completely sealed in stainless steel, to be delivered for local generation. When five years are up, the entire unit would be swapped out, the old unit taken away for maintenance and refueling.

I make no argument here about whether these technologies will be here soon, or are even technically feasible. I make no argument here about waste disposal or transit, although the travelling-wave reactor appears to address many of those issues before they arise. I am instead exploring the requirements of the smart grid with reliable power.

One of the oddest characteristics of energy markets today stems from the most reliable energy sources. Wholesale markets in energy regularly go negative for brief periods. If you have, say, a nuclear plant, you have a fire-hose of energy, one that cannot be scaled up or down easily or quickly. Too much energy on the grid at one time leads to spectacularly bad outcomes (except if you think explosions are pretty). If you operate such a plant, you pay other plants to go off-line during times of low demand.

So I began wondering, what if most of the energy on the grid came from such sources. What if no one was willing to accept an offer to reduce production because they too have difficulty scaling production up or down? How would we want to design the smart grid in those circumstances…

In this new reality, generators and grid operators would want to offer incentives for increasing energy use during off-peak hours. Smart homes and Buildings would want to install storage capabilities to transfer energy uses to these times when power was, at last, “to cheap to meter”. These end nodes would use this stored energy not only for their own use, but perhaps sell it back to meet the peak loads of their neighbors in this future time without expensive dirty peaking generators.

Buildings would choose different energy storage strategies based on storing energy for internal or external use. Thermal storage a can be a good solution for shifting load where heating or cooling is important. Kinetic, potential, or chemical storage might be better if the energy is intended for resale. Industry would want to consider, as it sometimes does now, changing its schedule to use energy when it is plentiful.

These use cases would require market operations to offer incentives for storage and resale to buildings and homes. We would need a model for time-sensitive prices to align use with scarcity and abundance. We would need symmetric negotiations to support returning of power to the grid without hierarchical control.

The smart grid requirements if we assume reliable power are much the same as if we don’t, and they would need participation from all parts of the market. Any node might be sometimes a supplier, sometimes a purchaser.

Standards Roadmap for the Smart Grid (SGIX) (v2)

Thanks for all those comments on my earlier post. I have updated the work and am re-posting.

The smart grid is more than improved top down control; it is a grid ready for unreliable energy sources (such as wind, waves, and sun), distributed generation, and Net Zero Energy (NZE) buildings. NZE buildings sometimes buy energy, sometimes sell energy, and energy use balances out over the day, season, or year. The NZE building presents particular problems as it may switch from buying energy one minute, and selling energy the next. Plug-in electric vehicles, whether hybrid or not, present the challenges similar to those of NZE buildings, with the added complexity of mobility. The smart grid requires distributed decision making, distributed responsibility for reliability, and easy interoperability to integrate an ever-changing mix of technologies.

The smart grid will be transactional, with each decision to buy or sell power a separate transaction at a separate price. The price of these transactions will vary dynamically, as a live energy market determines the clearing price at each moment for each sale or purchase. The smart grid will be open and transparent, wherein consumers can choose what kind of power to buy, and providers can prove that they are selling the kind of power they promise.

Alex Levinson of Lockheed Martin has named the suite of standards we will need for the smart grid as Smart Grid Information Exchange (SGIX). What follows is a personal view of a dynamic roadmap of the standards that comprise SGIX.

• SG Pricing: Price is more than a number. If I ask you if prices are up or down at the store, the answer is not “7”. It is not “Tomatoes are $3.00.” The price is “$3.57 per pound for the organic vine-ripened greenhouse heritage Cherokee tomatoes.” Each buyer can choose which attributes affect their purchase decision. I may choose to buy the cheapest tomatoes. I may choose to buy only organic. I may grudgingly choose the most expensive because they are the only ones in the store. SG Pricing will flow throughout the system—a model known as Prices to Devices. Under prices to devices, each system within a home or building may make its own decisions based upon budget and priority. I will be able to choose to run the fountain in front of my office only when wind power is available and below a certain price. SG-Pricing will be part of the SG Energy Market Information Exchange TC.
• SG Metering: This is a simple standard of energy flows by time slice. It also includes direction, as power may flow one way for a time, and then the other in a distributed world. To achieve transparent clearing markets, SG-Metering report what amount of what kind of power was purchased at what price at what time. If my neighbor and I buy the same amount of power at the same time, we may pay different prices because we may have made different decisions on how to buy. I may owe to my utility or to my neighbor for that purchase of solar power. SG-Transaction is in effect the accounting journal entry for each purchase or sale of energy.
• SG Energy Market Information Exchange: There is some bidding and exchange of information in advance. In my mind, this looks somewhat like commodity markets for those who want to participate. It includes elements of weather arbitrage. It includes time and reliability. It includes all of the elements of price. SG-EMIE will be developed in the Energy Market Information Exchange TC.
• UnitsML: UnitsML offers an unambiguous way to describe all physical measurements, and an unambiguous ability for a computer to look up the translation of any units of measure to any other units. SG-Pricing, SG-Transaction, and Energy Market Information Exchange will use UnitsML. UnitsML is an existing OASIS committee which will need some assistance and wider participation to complete.
• WS-Calendar: We all use ICALENDAR (IETF RFC 2445, http://www.ietf.org/rfc/rfc2445.txt) to unambiguously exchange information about time intervals. You used it the last time you clicked on an email attachment and suddenly had a meeting on your personal calendar. We need the same functionality standardized for web services. We will use it as part of pricing, and weather predictions, building management, and other decisions. WS-Calendar will be developed outside the SG effort as its anticipated uses extend into many business interactions.
• Digital Weather Markup Language (DWML): DWML is an existing specification developed by NOAA. NOAA offers a web service to which one can submit a longitude and latitude and receive in reply a DWML forecast. Most forward forward-looking energy markets are based on assumptions about weather. Most historical analysis of energy use includes recalling the weather environment. The most successful energy middleman base their business on understanding microclimates. We need to define a DWML profile for reporting as well as forecasting, to enable the exchange of actual conditions as well as forecasts. Such a profile would be used when querying local weather stations and even personal weather systems. Such a standard should include UnitsML (for internationalization) as well as time (WS-Calendar). We should encourage NOAA to develop the DWML specification into a standard; DWML also is of interest to the Emergency Response community.
• WS-DD and WS-DP: Device discovery and device profiles have been used in computer networking for some time. Device Discovery lets you find all printers on the network. Device profiles let you decide which printer to use when you want color duplexing. These functions are being standardized for the web. Schneider, one of the largest conglomerates providing systems for the grid and building is looking at providing WS DD and WS DP for all the equipment it sells. I think it will have a big role in the future world of distributed generation and Net Zero Energy facilities.
• SG Energy Interoperability: I envision this as a short, light, exchange of the information we need to plug technologies together without knowing the details. I see it as smaller than, but perhaps derived from, ISO-61850. It includes some basic safety information. It includes estimates of reliability and capacity. It may include some of the “price attributes” (Am I a source of carbon-credit eligible power?). SG Energy Interoperability includes critical Demand Response, i.e., non-market emergency curtailment of energy. A draft of the Energy Interoperability TC Charter is attached.
• SG-Load Control: The OASIS standard oBIX offers an extensible WS framework for communication with building control systems. OBIX defined a concept of Contracts, used to define higher level interactions. The ASHRAE BACnet Load Control Object offers a model for building systems to report on their energy use, to negotiate responsiveness, and to make load shedding agreements. SG-Load Control would build on the BACnet model to define a web service standards for contacts as defined by oBIX
• SG Telemetry: What is going on on the grid, and where is it failing. I recommend that we apply the watches, trends, and messages of oBIX into this critical area.
• SG Remote Operation: This one may be a literal transform from the ISO 61850 standard for substation communications. To the extent that SG Remote Operation moves into web services, it should apply interaction patterns and data models of oBIX.
• SG Curtailment. Sometimes, no matter how you plan, stuff happens. The daily temperature is 5 degrees warmer than expected. The turbine seizes. A truck drives into the transmission tower. Shed load NOW! Prices and markets for curtailment have been evolving rapidly; perhaps this addresses the grid integrity issues more directly. SG-Curtailment is part of the deliverable of the Energy Interoperability TC.
• SG Quality Of Service (SG-QOS): Participants in the smart grid must exchange information about reliability and performance. QOS information must be exchanged both as a promise and as a result. We may be able to adapt the Business Process QOS (BQOS) work from the EERP TC
• SG CyberSecurity: Security issues need to be integrated within every TC from the beginning—and not merely a veneer layered on after the fact. We need a separate security toolkit/framework, perhaps a profile from current fine-grained security standards, key management, and related areas. SG Telemetry may be an area to start defining so the broader integration of physical security, fine-grained networking and commercial security, and situation awareness technologies can be brought to bear.

Keep those comments and suggestions coming...

Why do we need all these smart meters?

Why do we need all these smart meters – so someone asked over at GreenTechMedia. We can run the grid with far fewer, and it will cost less. Why do we need these complicated protocols when we only need a price and a use? This perspective is correct; it is good engineering unencumbered with vision. This perspective is wrong; we cannot build tomorrow by doing what we day just a little bit better. Without pervasive metering, LEEDs and Green Buildings will remain a sham. Smart utility meters are only the first step.

Peter Drucker is still the most important and most visionary of thinkers about business and organizations. Drucker’s work ranged from identifying the long term sources of GM’s problems in 1942 to coining the term knowledge worker in the 1980’s. There are very few writers in this field whose work is more than a fad or a fashion. Few are worth re-reading. Drucker’s work is still relevant – even in the post DotCom world.. (http://www.amazon.com/s/ref=nb_ss_gw?url=search-alias%3Daps&field-keywords=Peter+Drucker).

Drucker fans are prone to invoking the pithy statements that sprinkled his work. My favorite Druckerism is "There is nothing quite so meaningless as doing well that which need not be done at all." There are many opportunities to invoke it at a state university where process often trumps any actual goals.

If we are creating a smart grid merely to meet the needs of the existing regulated market structures, much metering is not worthwhile. There are some limited benefits in peak management that accrue to the traditional utilities. There are few incentives for energy users to change, because benefits come to the diligent and free riders equally.  It is not worthwhile to have well-implemented smart meters everywhere if their interfaces are only for the power provider.

The real opportunity of the smart grid is its ability to work with more business models then the current top down reliable far-away power for dumb buildings and homes. The smart grid will support a network of power, with a network of new business opportunities for technology to insert itself into the energy chain.

Alternative Energy changes the grid because it is unreliable. If any significant amount of power on the grid comes from unreliable sources, we will have more peak energy events, when demand exceeds supply, per day than we now have per year. Distributed energy means that the neighborhood wind farm is now a full peer on the grid. Net Zero Energy means your dishwasher might bid against the grid for the output of your solar panel.

The smart grid offers choice. Homes and business will choose what power they buy, and they will want the smart grid to leave audit trails that they actually are getting it. A decade ago, supermarkets laughed at the idea that a significant number of consumers would choose more expensive groceries. Today, Whole Foods has transformed that industry and nearly every chain offers an organic produce section.

Why, you may even buy conventional reliable power to run your business but tell the fountain out front to run only when it can buy wave power. You may agree to pay a slight premium for your neighbors wind power when he is on vacation to keep his system working. We, or our software agents, will be active market participants in the national smart grid, in regional smart grids, in neighborhood smart grids, and even in in-building grids.

More metering, and more functional metering is worthwhile. Minimally functional metering is merely a way to reduce meter reading, not a step to the smart grid. And so, a final Druckerism: “We need to Measure, not Count.”