Power and Control
By M. Simon
Power and control. Every one wants it. The more the better. In America we have a lot of it. Some say per capita the most in the world. This may be true. Some say we may be the richest in the world because we use so much power. This also may be true. We do know that electric power usage is rising at a rate of 1 to 2% a year. Not too fast compared to what it used to be (about 3%) but still faster than population. So we either have to build more generating capacity or import electricity from one or two time zones over. Since demand peaks at different times each day in each time zone. You can use the spare capacity in one zone to support another. Because of the transmission distances and the losses such electricity costs more than it does at the generator but it still may be cheaper to use spare capacity instead of building new generators. Power is measured in watts or kilowatts, which is a thousand watts, very roughly one horsepower. Energy is measured in kilowatt hours. Power times time.
Next is control. We have a generator connected to a city on an island. We want to keep the lights on in that city. How do we control the generator? We make it run at a constant frequency (60 cycles a second in the USA) then we add a regulator to control the voltage. We make steam in a boiler and use it to run a steam turbine. The frequency regulator controls the amount of steam coming into the turbine. The voltage regulator maintains the output constant with the city load that varies over time depending on weather conditions time of day, day of the week, and scheduled holidays - Super Bowl Sunday for instance.
What happens when a big load is switched on? First the generator slows down and that means the frequency starts going lower. So the frequency regulator puts more steam into the turbine to keep the frequency constant. All the while the voltage regulator is doing it's best to keep the voltage constant. After a while things settle down and the city gets its steady power. In any control system you want the power output to follow the load demands as quickly as possible. The longer the delays the harder it is to keep the system under control.
Now let's take that island city and put it in the middle of Kansas. Let us connect it to a power grid that has a lot of connections to power plants in Iowa, Nebraska, and other cities and towns, and farms in Kansas and other places. One of the things that happens with this sort of setup is that all the generators operate in unison so frequency once established doesn't have to be adjusted (except for special circumstances). So now what you need to do is control the output power, because the frequency is controlled by the power grid.
Now the grid operates at the speed of light. With a 60 cycle frequency light travels 3,100 miles in the time it takes to complete one cycle. Control theory tells us that in a control system if there is more than a one quarter cycle delay between the control output and the controlled element the system will not be controllable. Now in power grid 775 miles is one quarter wave delay at 60 cycles. In actual fact because of the physical nature of the transmission lines the actual distance is more like 500 miles. This means in practice that at distances greater than 200 miles the system oscillates. Now we know this doesn't happen (noticeably most of the time) so how is this possible?
First we have a predicted demand curve. This helps to tell us when to bring on line or take off line various generators depending on their expense. The second thing that helps is that we have a lot of resistive loads like light bulbs which draw more power as the voltage goes up. Thus when a load is turned off the system voltage rises momentarily and the light bulbs absorb the energy and get a little brighter for a few seconds until the system adjusts. So the light bulbs damp the system. The other thing that helps is that most of the load is close. The delays are small and the loads are added and subtracted a little at a time. The system has time to adjust. And it will need to adjust on its own because to keep things from oscillating we are going to ignore all fast changes in power demand (since they are relatively small) and only respond to changes that last a while. So now no oscillations but the system is slow to respond and there can be small short term uncontrolled fluctuations as loads are added and subtracted.
Since the system was first built the nature of some of the loads has changed from resistive where the power goes up as the voltage goes up to a constant power type load (your computer for instance) where as the voltage goes up the current goes down to keep the power constant. This is called a negative resistance load. The more of these type loads on the system the less stable the system. Again what helps is keeping the load close to the generator. Reduce the delay.
Evidently what happened in the great blackout of 03 (and this is just a guess) is that a large amount of power was suddenly demanded in a section of the grid 500 miles from the generating capacity. Then just as suddenly the power started flowing in the opposite direction.
When a generator is over loaded it can trip off line to save the generator from overload. Generators are expensive and take a long time to replace. on the other hand if a generator loses it's load and there is steam coming into the turbine driving the generator the turbine can go too fast and fly apart so it will trip off if there is suddenly no load. Turbines are expensive and take a long time to replace. The same is true for the power lines.
So the power is surging back and forth in the system faster than the system can respond and there is not enough positive resistance load to absorb the transients. Generators start tripping off line from over speed or over load. This of course causes more transients too fast for the system to respond to and all of a sudden there is a cascading failure. The blackout of '03.
What can we do to prevent this from happening again soon? Today not much. Increasing the transmission line capacity is not the answer because it was uncontrolled power surging over long distances that caused the breakdown. More local generation will help. Match the local generation capacity with the local load to the greatest extent possible. Keep the fraction of power imported or exported from a region as low as practical. Encourage where economical more resistive loads. If power import/export is required over long distances use DC lines so the control problems are easier.
For the future what we need is local high power, high energy storage that looks like a resistive load when it is charging up and a local generator when power increases are called for. There is such a device and I have written about them here:
It is a super flywheel (almost sounds like a movie title, don't it?). They store a lot of energy per dollar. Unfortunately except for use in satellites they are still in development. In time they should see wide spread deployment. For one thing they could not only stabilize the grid but also double it's capacity because currently the grid is designed for peak usage and it does not run at peak most of the time. The problem with these devices is that we will not have them in megawatt type production quantities for probably 5 to 7 years. Thousand megawatt type units or farms will take at least another 5 to 7 years.
Long term what is needed is a strong development program to get prototypes out in the field as soon as possible. As soon as we find out what works scale them up. Once they are scaled up go into mass production.
Our problem it seems is not energy. It is power. The ability to deliver energy in a timely fashion. The timely fashion part is control. So what we want is more power and control. The sooner the better.
M. Simon is an industrial controls engineer for Space-Time Productions and a Free Market Green
(c) M. Simon - All rights reserved.
sierratimes.com |
