How renewables impact the grid

[AtrusBatleth 0]

Adding renewable energy (primarily wind and solar) for electric generation has some rather interesting consequences to our electric grid, and I think a real-world example might help to illustrate. First, for those who might not know much about the electric grid, you need to understand that the electricity you are using right now, is being generated right now. This very second. If you turn on a lightswitch, the power plant supplying your power needs to increase it's output immediately; if it doesn't, then the extra drain on the grid will reduce the line voltage (in electricity, voltage is like pressure in a city water pipe; open too many faucets and the system pressure goes down). Fortunately with a large enough grid, the extra power from the light you turned on will reduce the line voltage by a negligible amount, and also might coincidentally line up with another customer turning off an equally sized light at the same moment. But even with a large grid, the total power being consumed (called the demand) is changing with time. The larger the grid, the more slow and predictable that change is, but the demand still goes up during the day, usually peaks in the early evening when people are getting home from work, and then goes down at night.

Now for the real-world example. The attached picture shows a typical weekly demand forecast in the fall for a midwest electric grid. First notice the top line (dark blue). This is the total demand that needs to be supplied, in units of MW. The first day on this forecast is a Wednesday, and you'll notice the demand gets lower on Saturday and Sunday. It also gets lower every night, but it never gets below about 3500MW. This is called the "baseload." Before renewables came around, the grid was supplied by two types of plants: baseload plants, and peaker or load following plants. Baseload plants are designed to operate at a steady power level that optimizes the cost of operation (cheapest cost per MW). Peaker plants are designed to cycle on and off, ramping up and down in power to follow the changing demand curve. Nuclear and coal stations make for great baseload plants, while natural gas stations are ideal for peaker plants because they can ramp up and down power rapidly without any negative consequences to plant equipment. But for a natural gas plant to operate at steady power, it typically cannot compete with nuclear and coal in terms of cost per MW, which is why utilities would only use them as peaker plants.

Now enter renewables. The picture shows the forecast solar generation in yellow (it is pretty minimal in this region), and wind generation in green (much more significant). The red line represents the net demand, minus wind and solar generation. The red line is now what the nuclear, coal, and natural gas plants need to produce with renewables on the grid. Sometimes the wind is strong, such as in the early morning of 9/27. But that means the "baseload" is now reduced to about 1700MW. So what do you do with 3500MW of baseload generating capacity when the new baseload is only 1700MW? It means traditionally baseload power plants now need to reduce their output and become semi-peaker plants. This presents a number of problems for nuclear and coal plants. For one, they were designed to operate at a steady power, so ramping up and down frequently causes increased probability of equipment failures and maintenance costs. For two, there is some minimum power level that they cannot go lower than without shutting down completely, and once they shutdown it may take hours or even days to start back up again. For three, they were designed to optimize plant efficiency at rated full power, so whenever they operate below full power they are much less efficient. Operating less efficiently means the cost per MW goes up, making them less likely to be the utility's generator of choice.

This cost difference gets even more extreme when you consider that electricity is purchased and sold on a power market much like the stock market. Each power generator determines its cost of operation and submits a bid to supply a certain power for a certain price. The market takes the cheapest generators first, then the next cheapest, and so on until the forecasted demand is satisfied. The most expensive generator is what determines the price for electricity at that moment, and the ones that bid cheaper are therefore making profit. Renewables can bid extremely low, because their generation costs are artificially low due to a myriad of subsidies, some for initial construction and others that continue during the lifetime. So renewables are always preferred. So what if the forecast is so low at night (like the morning of 9/27) but will go back up again in the morning? Of course the natural gas plants can shutdown easily, but with the demand so low some of the nuclear and coal plants will also need to shutdown. But then they wouldn't be able to startup in time for the increased demand in the morning, so they keep operating. In order to have somewhere to send the electricity they're making, they actually PAY renewable generators to stop generating. So in the early morning of 9/27, those nuclear and coal plants are taking a loss because they are paying more to windmill generators than they make for selling their power. Even with this arrangement (which windmill operators love), if the utility forces windmills to not run too often, they actually get fined and penalized by the regulator for not utilizing the wind generation that stakeholders invested in. This is in addition to the continuing subsidies per MW that renewable generators get. So there is a very strong financial incentive for utilities to maximize generation from installed renewable capacity, even if demand is low.

But we're adding renewables to the grid, so doesn't that mean we can start permantnely shutting down those non-renewable plants as no longer needed? Isn't that the whole goal of going renewable? Unfortunately, it doesn't work out that way. Because renewables are less reliable, less consistent. For example, on this weekly forecast wind is generating over half of the demand in the early morning of 9/27. Hurray for wind! But look at the evening peak on 10/1. The wind is pretty low, right when the demand is highest, so almost all of that peak demand (about 5000MW) needs to be supplied from non-renewable sources. The solution most utilities have pursued is to replace aging baseload plants that can't cycle power very efficiently with natural gas plants that can. This way they take the renewable generation when the weather cooperates, but can rely on natural gas as backup for when the weather doesn't. Remember that natural gas plants make great peaker plants. They can ramp up and down rapidly, shutdown completely and startup again within minutes, and their efficiency is about the same regardless of what power level they're opeating at.

This is why you see such an increase in natural gas plants. They produce less CO2 per MW than coal, that's true, but that's not really the driving force for replacing coal with natural gas. It's because utilities now need peaker plants that can deliver nearly the full grid capacity to serve as backup for inconsistent renewables. But for now, this system works. It helps that fracking has drastically increased the economically recoverable reserves of natural gas. If not for abundantly cheap natural gas, this grid system would cost a lot more to operate. And nuclear and coal plants, now being forced to operate less efficiently, are having a hard time competing with cheap natural gas and are therefore shutting down. But we are far more dependent on natural gas now than we used to be. Not just the extraction of natural gas, but its delivery via rail or pipeline. This added reliance has already resulted in some supply disruptions during midwest winters, when on extremely cold days the natural gas for heating homes was prioritized (rightly so) and there wasn't enough supply left over to run all the natural gas plants which had to reduce output (luckily there were still enough nuclear and coal plants available to pickup the slack during these times).

While this example focuses more on the effect of wind generation, solar generation has it's own interesting effects. Solar generation peaks in the middle of the day, steadily getting lower as the sun lowers. But the demand in the attached example stays consistently high throughout the day (at least on weekdays), and peaks in the early evening just as solar generation is reaching near zero. This makes the net demand (red line in the attached picture) get low during the middle of the day, but then rapidly skyrocket to it's maximum during the evening peak. This is commonly referred to as the "duck curve" (because apparently the curve resembles the shape of a duck, but I don't really see it). For a grid with a large amount of solar generation (which is not the attached example), this also favors adding natural gas capacity to backup the solar generation. Natural gas wins again!

Speakers corner is for us to get on our soapbox, so that's what I've done. But I do sincerely hope some of you reading this have a better appreciation now for how the grid operates, and maybe even are interested enough to do some further research of your own. If you want to learn more about this, I highly recommend you lookup your local independent system operator (ISO) and poke around their website. There's MISO in the central region, CAISO in California, NYISO in New York, etc. Contact them and ask for a tour of their control center. If you think I'm just blowing hot air and wasted your time, well it's just my soapbox and you can keep walking by pretty easily. Blue skies.

Max Peck
What's the point of having top secret code names, fellas, if we ain't gonna use 'em?

[ryoder 1,590]

I'm seeing something missing from this discussion: Grid energy storage.

https://en.wikipedia.org/wiki/Grid_energy_storage

(Note that the water backed up behind dams is already a form of storage.)


Storage is part of my local utilities plans:

Under Xcel’s Colorado Clean Energy Plan (CEP), the Comanche coal units will be replaced with a $2.5 billion investment in renewables and battery storage — including of 1,131 megawatts of wind, 707 megawatts of solar PV, and 275 megawatts of battery storage across the state, including in Pueblo. Xcel estimates the transition will save ratepayers between $213 million and $374 million.

https://www.greentechmedia.com/articles/read/xcel-retire-coal-renewable-energy-storage


Australia already has battery storage online:

The Tesla lithium-ion battery in South Australia is on track to make back a third of its construction costs in its first year of operation, new financial documents show.

https://www.theguardian.com/technology/2018/sep/27/south-australias-tesla-battery-on-track-to-make-back-a-third-of-cost-in-a-year

"There are only three things of value: younger women, faster airplanes, and bigger crocodiles" - Arthur Jones.

[AtrusBatleth 0]

Good point, I had intended to mention battery and pumped storage but forgot. That is another solution being pursued, but up until now it has been very challenging and cost-prohibitive to implement on a large scale. Plenty of off-grid homes have their own battery banks, and they think "why can't utilities just do that?". They have no idea just how much power the grid is consuming. Most of the demand is not even residential, but commercial. There would be some pretty severe consequences from implementing that much battery production and disposal. But it's a developing technology and I'll be curious to see how well it actually lives up to the hype. Especially in terms of real cost.

Max Peck
What's the point of having top secret code names, fellas, if we ain't gonna use 'em?

[AtrusBatleth 0]

Colorado CEP article

Not all of the clean energy projects pitched in the CEP are needed by the system at present, Xcel noted. However, the utility argued that these projects offer immediate economic and environmental benefits, because PSCo will be able to take full advantage of federal tax credits for wind and solar projects before they expire, allowing the utility to deploy a greater amount of renewable energy resources than it could justify otherwise.

Not arguing in favor or against, but just pointing out that the current renewable construction projects are still reliant on government subsidies. Xcel Energy does a phenomenal job with clean energy though, making real progress towards the lofty goal of 85% carbon-free by 2030. Of course part of that 85% includes their nuclear units which will reach the end of their current license in 2030 and the few years following. So unless they renew the licenses (which I hope they do, but there are those in the MN-PUC that don't even want them to operate to 2030), they will quickly fall below that goal just after having achieved it.

Max Peck
What's the point of having top secret code names, fellas, if we ain't gonna use 'em?

[billvon 3,118]

Quote

Not arguing in favor or against, but just pointing out that the current renewable construction projects are still reliant on government subsidies.

In some places yes, in some places no. In Saudi Arabia, large solar projects are going in without subsidies. The Jinko/Marubeni project, for example, is putting 350 megawatts near Abu Dhabi with no subsidies - and for a cost of 2.42 cents a kwhr.

[AtrusBatleth 0]

billvon

Quote

Not arguing in favor or against, but just pointing out that the current renewable construction projects are still reliant on government subsidies.

In some places yes, in some places no. In Saudi Arabia, large solar projects are going in without subsidies. The Jinko/Marubeni project, for example, is putting 350 megawatts near Abu Dhabi with no subsidies - and for a cost of 2.42 cents a kwhr.

Good point of clarification: I am focused on the US grid. Other countries definitely vary.

Max Peck
What's the point of having top secret code names, fellas, if we ain't gonna use 'em?