James C. Liu's Weblog

Friday September 10, 2004

Hot Weather Power Crunch

It's finally been cooling off after several warm days in California. At least I can say that because I'm in the Bay Area and we enjoy milder conditions than those folks who live inland. Summer is late this year, but finally here. Along with that hot weather were a couple of incidents this past week where the California Independent Systems Operators (California ISO) issued warnings to the public to conserve power between 4 - 7 pm. The high temperatures inland were creating a huge demand for electricity to power air conditioning and refrigeration units.

The reasons stated by most news agencies and the California ISO folks is that electricity demand has been rising and California state utilities have not constructed new power plants sufficient to fulfill the demand. This may be a relatively new revelation to some CA residents who've not experienced such shortages until the last 4 years. But I have to say that about 20 years ago, when I took my first Nuclear Engineering class at Berkeley, one of the first charts thrown up on an overhead projector was the rise in electricity demand relationship with GDP growth. The correlation is well known and understood by the Utilities and their watch dog Public Utilities Commission (PUC) organizations. In fact, managing the electrical generation capacity is also well understood.

In general, a large electrical utility has some constraints on how the electricity must be delivered. Usually there is some ANSI standard that defines such qualities, such as the AC power needs to be 60 Hz in frequency with no more than a small fractional change plus or minus, and the voltage sent over transmission must not vary by more than 10% either way. So ideally, Utilities strive to provide clean 60 Hz power at 110 VAC into most homes and businesses (single phase power). But the other challenge here is that power is not stored significantly in the grid. It must be manufactured pretty much at the moment that people use it. For a small household with a diesel or gas generator, when someone eats more power, the engine speeds up and burns more fuel and increases power to meet the need. This is no different. However, because of the large variation in power quality from a small generator, there's usually quite a bit of power conditioning circuitry built into the device so that it pre-conditions the power to meet the needs of electrical appliances. But such power conditioning does introduce some reduction in efficiency and loss of power, so we really don't want to do that much. This is also the issue with wind and solar generators. We need to condition the power they generate to be of sufficient quality and furthermore, if they are connected to our grid, we want to make sure they stay in synch with the rest of the power grid or else they may introduce some nasty harmonics that should cause problems with the main grid.

One way is to have a small resevoir for power, like a battery, or if we recall the Popular Science magazine articles back in the 70's, we could have underground, vacuum chambers with mag-lev kinetic energy coils spinning at 100,000 rpms to store power. Only, I never could understand how the kinetic energy thingamajigs would continue to spin at 100K rpms once we started to suck the power out again. Wouldn't we still need inefficient power-conditioning devices?

Well, yes, we'd still need power conditioning devices. And that's why large utilities are still in business. While they are big Monopolies that appear to have an evil ability to extort a monthly subscription fee, in fact, they are still a lower, and hopefully more efficient and environmental way to produce power rather than have millions of users in every locale running their own generators and storing them in chemically toxic batteries or explosively deadly hydrogen containers. Modern power plants are able to burn fuel to heat water to steam and extract work at much higher temperatures in turbines than any smaller commercial or private facility using diesel or gasoline generators. And this is governed by the maximum limits as defined by the Carnot Cycle (even though steam turbines operate on the Rankine cycle).

Big power plants also offer better quality of power because they aggregate power demand into a statistically consistent electrical load - also referred to sometimes as "baseload." A single household might go from zero kiloWatts to 4 kiloWatts power consumption in 2 seconds, just by someone cooking dinner and some teenaged child blow drying her hair in preparation for a Friday night out on the town. But aggregated over say 500,000 households, the demand looks pretty consistent, and any one household's fluctuations will have infinitesimal impact on the frequency at which the turbines spin, just a miniscule short term voltage drop which the rest of the grid easily absorbs (which is why the lights dim and flicker slightly when the microwave goes on but recovers within a fraction of a second).

Typically, a utility has some historical data. Based on the time of day, day of the week, month of the year and current weather conditions, they can predict the power demand and hopefully meet all the needs. And to do so, they use some simple rules. Based on GDP growth, they can predict the long term demand and build out more power plants. Based on statistics on maintenance and operations, they also know how much excess capacity in power plants they need to keep available because at any moment, they might have a plant down for maintenance and service or usually warm or cold weather may increase demand for electric power. For most utilities, the magic number falls somewhere between 17% and 20% excess capacity OVER average PEAK load. Which means, take the biggest demand for power (in California, it was set just this Sept. 8th at over 45GigaWatts of demand). We had warnings issued Wednesday to turn off lights and shutdown unneeded systems inside our company. And it was a good thing to do. California only has about 2.5 GWatts of excess capacity, which is only about 5% - 6% over. A single shutdown at say, Unit 1 at Pacific Gas and Electric (PG&E's) Diablo Canyon Nuclear Power Station near San Lius Obispo could lose 1.1 GWatts-electric for the state and cause large shutdowns in many critical sections of the state's grids. A small but survivable 6.5 earthquake near the Central Coast could require both reactors to SCRAM and shutdown, leaving just a 1% margin in power. So typically it is wise to have a safe margin like 17 - 20%.

There are some ways we could offset some of the demand, especially in sunny California. We could require Photo-voltaics to be standard installations as roof tops on all new homes built after 2010. We could start today and mandate, say 10% of new homes or rennovations exceeding 60% of the homes value to include at least 1 kWatt of photo-voltaics. Some folks argue that photo-voltaics are actually quite unenvironmental because the energy required to make them may never be recovered in the expected lifetime they will be used for. There is some truth to that, but research by at least several folks at UC Berkeley and the Lawrence Berkeley National Lab may be changing the power balance equations. There is also the truth that solar panels are DC generators and we want AC. We need an inverter and thus it's less efficient. Still others suggest we forget the whole scheme and simply replace old outdated fossil fired plants, because like old cars, they're the big polluters and very inefficient. A newer plant burns cleaner and has a higher Carnot efficiency and is therefore better for the environment.

Personally, I think such discussions are all good. Perhaps we -should- replace old, inefficient power plants with new ones. We may even want to invest in cleaner power that doesn't put CO2 in the air, like nuclear fission plants, or maybe in 25 years, Inertial Confinement Fusion (mini-Deuterium-Tritium bombs shot into a chamber and targetted with lasers or actually heavy Ion beams at 8 times per second - c.f. HYLIFE-II - the High Yield Lithium Injection Fusion Energy design - LLNL 1993).

In the meantime, I wonder why we aren't seeking renewable alternatives, more. There are products that some entrepreneurial business guys could sell. For example, when the heck will I be able to buy a self-installable solar water heater from Home Depot or Lowe's? When will I see a weight-loss program that is actually a bicycle hooked up to a generator that you plug your TV into? If you wanna watch TV, you gotta pedal! 3 easy payments of just $59.99. Results guaranteed! And When are Utilities going to start installing power monitoring and harmonics arrestors plus a power shunt on every home power line to allow, say, upto 10 kWatts of independent power production from Wind, Solar, Kids on a bicycle generator, etc. It may not work in other parts of the world where it's hotter, muggier, still, and cloudy. But in sunny California (and some places breezy), we could have 2 or 3 GWatts of capacity by year 2010 installed in these home based systems that have 80% availability from 1pm - 6 pm most days and that could save tonnes of CO2 put into the air, and help provide more safety margin in our grid.

I'm not sure why we don't have these things. Maybe society is waiting for Java/Jini/JXTA to be installed on these energy appliances. I did work with a midwest Energy Utility before on embedded commercial air conditioning controllers that had a JVM and network stack. It's been a while since I talked to those folks. I'm not sure if anything ever panned out on their product side. Maybe quite a few of us are more focused on conservation and not looking at increasing alternative production. I'm not sure. I do know that I should conserve, and I should get back on my bike too. I might not save the planet with those actions, but I might just get healthier and save myself. September 10, 2004 05:41 PM PDT Permalink