Lets start by talking about the electric grid sense understand it is important for understanding the issues with intermittency. With the electric grids the amount of electric power produced always needs to equal the amount used. Matching production and use with uncontrollable and difficult to predict sources like wind and solar can be tricky. Things like clouds and changes in wind speed can cause problems. One way of understanding this problem is to think of the electric like a giant bucket.
"The Western Grid is like a giant bucket," said Mark Avery, SRP's grid manager "with a bunch of spouts running in and out, and you have to keep the water level constant." The Denver Post
Picturing the electric grid as a giant bucket. Some people are taking cups of water (electric power) and pouring them into the bucket while others are taking cups of water out of the bucket. If the bucket become empty it’s bad because people can’t get their water, and it’s also bad if the bucket gets too much water and starts overflowing. The water in the bucket isn't very deep (just enough for someone to get a cup full) so the rate of the water going into the bucket has to precisely match the rate of the water coming out of the bucket. If there is only one person drawing water from the bucket this can be difficult to do. One person is fairly unpredictable. What if he all the sudden decides he wants a lot of water, or what if he all the sudden decides he doesn't need any for a while. This makes load following (i.e. making sure the right level of water is always present) more difficult and less efficient. Lucky the actions of a lot of people average out into something much easier to predict. So in order to deal with the problem they made the buck wider (but still just as deep) so many people can draw out their cups of water at once.
This system worked well enough (most of the time). Then one day some new people (i.e. wind and solar advocates) decided that they wanted to put their cups of water into the bucket as well, but other people didn't want them to because they couldn't control when they put the water into the bucket, and because they also couldn't predict it with perfect accuracy. The new people said it would be fine, and that just like with people taking water out of the bucket things would become more predictable if they just made the buck wider so more people could put their cups of water into the bucket at once. Then once things became predictable the people that could control the rate they put water into the bucket would help match everything up.
So how well does this new way work? Opinions vary, but personally I am very sceptical that adding different types of unpredictability together will somehow make things more manageable. One thing’s for certain, the electric grid is not really a bucket. It is an expensive complex machine, and making it do what the renewable energy advocates want makes it even more complex and expensive. I think that’s why they are always saying things like “we need to upgrade our archaic electric grid” or “we need a smart grid”. Sure the electric grid (just like roads) needs maintenance, occasion expansions and even upgrades; but I believe that the biggest reason they are pushing so hard is because they want the money needed to integrate more solar pv and wind without having to included that money in the costs of those technologies.
So what happens when things don't match up? Well larger difference cause Power outages while smaller differences cause other power quality issues. Both of these things have costs. A Berkeley Lab Study estimates that power interruptions cost the US $80 Billion annually.
Lets talk a little bit more about power quality. What is power quality? Opinions vary but here is one definition I found useful.
"Power quality is simply the interaction of electrical power with electrical equipment. If electrical equipment operates correctly and reliably without being damaged or stressed, we would say that the electrical power is of good quality. On the other hand, if the electrical equipment malfunctions, is unreliable, or is damaged during normal usage, we would suspect that the power quality is poor."
We have standard for voltage, frequency and phase. Then we make devices that run off those standards. If the power difference to much from the standard then devices won't work properly or they can even be damaged. Both Solar pv and wind can cause power quality issues (e.g. can deregulate line voltages and sometimes in extreme circumstances even shifting the line phase ). Google the words wind and solar along with power quality and you can learn about the various issues and proposed solutions, or you can watch this video (I highly recommended it). There are sighs Germany is already having problems with it’s level of penetration.
"short interruptions in the grid has increased by 29 per cent in the past three years – resulting in some firms on the grid reporting damage running into hundreds of thousands of euros as a result of unexpected stoppages."Manufacturing requires good power quality which solar/wind can have trouble supplying. This is especially true for manufacturing high tech things like solar panels. There have been attempts to deal with the problem with things like battery back up at the source, but there are still signs that large amounts of wind and solar can cause problems. In order to cope with these problems manufactures need to spend money on special systems (for example system that use battery backup), but such things have costs. However the problem is dealt with (e.g. at the source, smart grids and/or making the end users deal with it) there are costs that should be included in the price of wind and solar.
The variability of wind and solar means that other types of energy generation have to ramp up and down more often in order to match electric production with use. This creates inefficiencies which have costs that should be attributed to wind and solar.
A good way to understand these inefficiencies is to compare electric generation to something most people are familiar with. Cars are more efficient when they are driven a certain way. For example.
"While each vehicle reaches its optimal fuel economy at a different speed (or range of speeds), gas mileage usually decreases rapidly at speeds above 50 mph."Power plants also have an optimal fuel economy when operated at a certain continuous output. They call plants made to operate at their optimal fuel economy Base load Power Plants and anything that causes them to very from their continuous optimal output creates inefficiencies that have costs. Some of that cost should be attributed to wind and solar (The rest of it should be attributed to things like changing demand).
It's important to note that power plant not operating at their optimal output because they are being used for load following(i.e. being used to help match electric production with use) are performing a service for the gird. This service is called spinning reserve and studies have been conducted to try and estimate how much it costs. One such study is quoted below.
"An expected finding from case studies made to date is that the specific cost of power generated in spinning reserve mode is quite high compared to the optimum cost of power from the same unit. This is, of course, due to the poor heat rate of most thermal power units at low load. If the unit could have operated at high load instead of spinning reserve, there is a lost opportunity cost which may double the cost of the spinning reserve service."
Next is another comparison between cars and power plants.
"Idling can use a quarter to a half gallon of fuel per hour, depending on engine size and air conditioner (AC) use. Turn off your engine when your vehicle is parked. It only takes a few seconds worth of fuel to restart your vehicle. Turning your engine on and off excessively, however, may increase starter wear."
Unlike internal combustion engines base load power plants can't start up that easily (Some can take more than 12 hours to reach full load). How long it takes to start a base load power plant varies based on numerous factors. One such factor is how hot it is. Cold starts take the longest while warm and hot starts take less time. Trying to get the plant online too fast can result in unnecessary plant failure or wear. This bring us to another cost that is increased by intermittency. Intermittency increases Power Plant Cycling Costs. Power Plant Cycling Costs are the increased costs of maintenance and forced outages caused by things like turning the plant on/off, load following, and minimum load operation, in response to changes in system load requirements. There are ways to reduce these costs like keeping the plants hot, but such things also have costs.
Another costs of intermittency is as the cost of underutilized capital assets. A good example of an underutilized capital asset would be a power plant that only runs a few month out of the year when its too cloudy for solar pv, or a transmission line going to a wind farm that has to be build to handle that wind farm’s maximum capacity even though on average the wind farm only delivers 30 percent of that. Here is a good example of the problem from Germany.
As you can see there are days in January with almost no wind or solar production. The question you might be asking yourself is how do they get power when wind and solar aren't there for them. What happens is that people end up having to have two power systems. The conventional power system (mostly coal in Germany) which is able to meat all of the countries needs plus an extra wind and solar system which can't be relied upon. Both of these systems have to be paid for which as you can imagination is quit costly. A lot of people seem to think that some costs don’t count, but if people want to continue to enjoy electricity on demand 24/7/365 then they do count and they need to get paid.
In conclusion there are reasons why electric prices are higher in places that embrace solar and wind. The sticker price they show you isn't even close to all that you'll have to fork out. This shouldn't be allowed to go on. There need to be a better accounting of the true costs of producing electricity with different methods. Some people have already started on it, but a lot more work need to be done.
Update Mar 4 2015: Made some changes on things I didn't like.