The Argument for Market change: Case Closed.

Friday, Lynne Kiesling blogged about the problems illustrated by the New York blackout last Wednesday (http://www.knowledgeproblem.com/archives/002106.html). I repeated some of my comments about generally bad power and the inability of the current market to address them.

Another poster, Ed Reid replied (I abbreviate, but the link above will let you see the original)

My neighbor, Toby, has different concerns about reliability than the typical regulatory commission, including North Carolina's. Most regulators are concerned about cumulative outages of less than one hour per year (99.98% reliability). The outages he refers to here are mostly either momentary "reclosure events" or local outages resulting from vehicle accidents. I suspect some electric utilities would be interested in selling reliability of the type Toby is concerned about, but I suspect very few customers would be interested in buying it, both because of the extremely high cost and the relatively modest inconvenience associated with these outages.

Ed is clearly knowledgeable about the current grid, but let me re-phrase this.

“The quality you seek is not what the Utilities Commission does – be content with what they have decided is good quality. The current architecture is not really able to give you the power modern electronic systems expect. If we do things the same old way, it will be too expensive to fix. It will only be offered if state commissioners decide there is a market for it.”

I could not possibly have made a better case for market change, for de-regulation, and for the benefits of switching to a market that offers incentives for innovation.

Making Decisions when Everyone Lies

On Wednesday afternoon around 3:45 the power went out in areas of Manhattan and the Bronx in New York City. Although the outage lasted less than one hour, it was sufficient to cause traffic congestion and inconvenience to almost 400,000 customers on a hot, humid summer afternoon.

Or so the official story goes. There was a one hour problem. The problem was gone in an hour. There was some inconvenience. And, as Dorothy Parker famously concluded “…And I am Marie of Roumania

In the 70’s, the American economy suffered through structural problems caused by more than a decade of high inflation. The effects of this inflation were a number on the nightly news. If it were a slow news week, a grocery manager and a housewife would be interviewed, with some statement that consumers were switching from steaks to ground beef. Well, I grew up as one of 11 children – we always ate a lot of ground beef.

The official line was there is a small problem. We can simply quantify it. There are some small concessions that must be made at the grocery store. All fine until the house of cards fell, and all the economic decisions based upon continued inflation compounded, and all the financial games to beat the system by sneaking around the facts came together in a no-job economy with 20% mortgages…

An essay read as I completed High School, one that I cannot find now, struck me. It was so clear, and predicted so much, that I will try to recreate its theme:

“Living with inflation is living in a country where everyone lies about all the most important facts. Every time money is borrowed, both parties are lying about what it will cost to pay it back. Every time someone sells an asset, every home, every factory, every business, they lie to themselves about whether they made money Every investment is made not to increase productivity, not to find value, but to game the lies that everyone tells”

Welcome to WIN buttons and cardigans, and to gas lines and the misery index.

Today we lie about power. We lie about its quality and availability. We lie about the cost of outages. Reliability cannot be purchased on normal markets at any price, because it’s not for sale. We lie about the long term costs of power incidents, and the long term destruction of assets and their resulting accelerated depreciation.

Friday I was checking up on the dorm opening date to drop my daughter off at NYU. The University had an alert on the front page:

Thursday, June 28, 2007, 5:30 pm
Power has been restored to the Fairchild Building. Based on our conversations with Con Edison, we expect normal power operations at the building tomorrow. All employees should report as usual, and operations and activities should go forward as scheduled.

Hmmm – so a one hour outage took the building off line for the entire next day. Another lie about the costs of the outage.

Such outages also cause long-term damage to capital assets. Across Manhattan, Compressors on air conditioners and refrigerators will fail early because of the outage, causing unmeasured expense and loss of service to office, restaurants, real estate owners, and residents. Water systems in high rises already suffered one pumping failure, and will again when the motor fails early, causing turbidity, and cases of minor illness. These illnesses, like the failures, will occur months from now and never be associated with this outage.

We have the most fabulous information system ever devised; the flows of money in an open economy. When piles of cash start getting clogged up in portions of the economy, innovators compete to get it flowing freely again. When we tolerate a closed economy, we create poor information and poor allocation of resources. We externalize costs in ways that we do not recognize. We make bad decisions.

We have a closed regulated energy market wherein the true costs of poor quality power, and even how poor that quality is, is hidden behind opaque markets run by utilities commissions. We do not know the true costs of the energy supply, in particular, how it varies with scarcity across the day. And because the energy market sits just outside our vision, invisible and uncontrollable, we let ourselves believe that it is of good quality.

If we knew the costs of power throughout the day, we would demand knowledge of its quality as well. If the quality were to become visible, we would begin to recognize the correlations, not only of big events but also of small events to PC failures, and to window AC unit failures a month after a poor quality power incident.

If we knew the costs, we would wonder why it isn’t cheaper to fix it. If we knew the true size of the costs, the fixes would not seem so expensive. If we were willing to buy, the innovations that will make real differences in power consumption, in power reliability, in cost of living, and even in carbon use would come to market.

But we have chosen closed markets in power. We have chosen markets wherein the important information is hidden from the consumer. We have chosen markets without instant access to true pricing and true quality. We have chosen markets based on half truths. And half truths are lies.

And because we lie, we make poor investment decisions, and it appears to cost too much to fix.

The Big Picture on Energy and Carbon

Demand Limiting. Load Control. These approaches to saving energy have been around in some form for a while. The problem is, the applications to-date have been centrally controlled. Sign up for this program months in advance, and then the Power Company controls your [water heater]. But what if I live in Marin County and feel that my hot tub / water heater is mission critical from Friday noon through Sunday? Too bad. Consumers do not want loss of control. This limits participation.

Recently, a professionally Green university administrator confessed to me that he “used to do that, but didn’t anymore.” What was the problem? For the few weeks a year that his kids came home from college, he could never get adequate hot water. There was no easy way for him to bow out of the program on short notice. It took phone calls, forms, lead times…so now he does not do it at all.

New initiatives are getting closer to changing this. There are now a couple web services protocols for building controls: oBIX and BACnet-WS. The newest Windows now is able to discover such services automatically using WS-Device. Soon there will be software to let your PC discover and operate building systems much as they discover printers today. It is not hard to imagine an agent talking to the power spot market, talking down to the systems, reading the electric meter live…

Some of the so-called “Zero Energy” initiatives envision each building supported by multiple on-site energy collection and generation systems. Based upon the building’s operating posture, and the mix of energy sources available, such a building would pull 35% or less of its total energy budget from the grid. If the facility includes local buffering and storage of electrical energy, whether this buffering is souped-up traditional batteries or new-fangled hydrogen storage, this becomes viable. On-site DC power generation can be stored in those batteries without the losses you would expect converting AC to DC first. If the building negotiates with the grid for spot pricing to influence the internal decisions, then it is also part of GridWise.

If future houses support DC distribution and internal use, then the batteries can become the primary source for the house. This increases the life of the batteries with no new storage technology, as it does away with the losses from converting back to AC. Most devices in modern houses are DC anyway, with very inefficient brick transformers that may convert a third of the power coming into your house to heat. The Galvin Electricity Initiative (www.galvinpower.org) is a good source for the engineering behind this. With appropriate local buffering, an awful lot of power consumption can be shifted to off hours without loss of occupant autonomy.

GridWise envisions a future power grid broken down into separate clearing markets for Generation, Transmission, Distribution, and Customer Face. Customers would be able to negotiate directly for green power at a premium price if they so desired. Others would be able to negotiate for the cheapest solution. Neighborhoods could opt for their own intelligent distributions systems, with higher reliability profiles than the grid provides. Customer Faces would aggregate customers with similar profiles, based on price sensitivity or reliability or social consciousness, for billing and for operation of home control systems.

The market oriented approach of GridWise is rare among energy initiatives because it anticipates driving efficiencies through heterogeneity consumer choice and thereby driving market innovations.

As the grid becomes “the thing that charges your batteries”, it becomes cheaper to put alternative energies on the grid. The grid does not need to be so concerned with frequency regulation and other arcana. This means “unreliable” sources such as solar and wind can increase their proportion of the energy load without taking down the grid.

All these innovative strategies are driven by more efficient clearing of actual instantaneous energy pricing, and making it less onerous for the consumer, whether home or business, to participate. The effect on power with today’s technologies and market structures, as some have said above, is limited. Hourly, or better minute-by-minute, energy pricing is a necessary precursor to developing the markets that make these strategies worth pursuing.

Still Waiting. . .(part 3)

 So if the control companies are now using XML, why am I still waiting?

All of the current web services standard offered by the mainline controls companies are REST, meaning they let me push a point in using XML, and they let me read a point using XML. To some extent, they let me get many points using XML, as I can request a log, or a trend, or a history, and get back many time-stamped values.

But none of them is abstract. Which means none of them feel like a printer driver. If you want, you can find a networked printer. Whatever kind of printer you find, you can usually print to it without further thought. Some of them are in color. Some of them allow double sided printing. Some of them have more than one bin. You can find out by inspecting it from afar. In any case, if what you want to do is print, you can just print.

You do not have to know how a printer works. You do not have to know how a fuser works. You do not have to know how to operate the ink jets. All those internal relays, they mean nothing to you. You can just print.

Contrast this to our general experience with engineered systems. Generally, you will not know what it is. Is it a thumb-drive, is it a camera, or is it a printer. Is it an air conditioning system, a power system, or an intrusion detection system? The system won’t tell you.

You will not know where it is. You cannot link it to Google earth. You cannot link it to your administrative space assignment. You probably can’t even map it to the CAD plans that were delivered to you with the building two years ago. You are expected to know.

Until these systems know what they are, you can’t really perform the most basic interactions with them. I would like to invite the A/C to the meeting, just as I invite other people. But what am I inviting? Is it the air handler above the ceiling? Is it the thermostat? Is it the outside air temperature outside that I cannot set at all? Until systems can describe what they are, I can’t use them. The enterprise can’t use them.

Until the systems know what they are, I can have no security. Well, I can have one security: keep away! I cannot say that all administrative assistants at the departmental level have the right to set the thermostat. I cannot say that maintenance personnel get to see deep details until the system can tell the difference between the deep details and the superficial. Without nuance, there is no security.

I have no way to integrate building systems into long running processes. They do not know what services they provide; they only have settings, and measurements. With REST, they do not understand work flow languages like BPEL (Business Process Execution Language). I cannot integrate them into management frameworks like WSDM (WS – Distributed Management) or like WS-Man (WS – Management).

Two weeks ago, I listened as a major building systems company asked the Data Center group at IBM what they wanted revealed. The answer “We do not want raw data. We want the best information that you can provide with your deep domain knowledge to present to us actionable information. We do not want to become electrical engineers.”

Well, that’s what I want, too. And I am still waiting.