A few days ago a postcard showed up in my mailbox advising me of an impending deadline for an energy storage incentive for households with installed solar arrays. Interested, but assuming it was a scam of some nature, I did some research and found that it was a real program. Put simply, the state and local utilities have decided that the model of distributed solar generation and storage, already widely used in Europe and Asia, is indeed something worth looking into. This was something I had wanted to incorporate into my system from the start, but when I decided to install the solar array there was no incentive program and costs were excessive, the optimized hardware had not yet been certified, and the utility was, quite frankly, dead set against the idea.
The buzzword at the moment surrounding this approach is “Smartgrid” and there are numerous ways it is being considered and implemented. At it’s most basic in this context, the idea is that the rapidly increasing availability of the “Internet of Things” (IoT) can be used to improve the efficiency and effectiveness of the electricity network. Using my solar installation as an example of a typical grid-tied, net metering system, the system works in a basic mode. Assuming that all is normal with the grid, when the sun is shining my panels generate as much electricity as they can. This electricity is then basically dumped into the power grid between my house and the utility meter. If my house needs all of that electricity at that moment it goes into the house; if not any excess goes into the grid. Likewise, if my house needs more electricity than the panels are producing, the excess is supplied by the grid. Per the terms of my net metering agreement (and indeed, what should be the terms for all end-user interconnect agreements) it is a 1 for 1 trade: for every kilowatt hour I give to the grid, I can pull one at a later time at no net cost. This can easily be seen as using the grid as a “battery” backup for the nominal cost of the minimum monthly fee charged by the utility to help offset the cost of the distribution network, and at the top level it works well.
The challenge in this system is that as more solar generation is added to network as a whole a cyclical generation cycle develops – all the solar rolls in during the day and out in the evening. The grid has to anticipate and react to this effect in order to maintain a steady supply to all users, and becomes yet more complicated as wind energy is included into the mixture. This drives a very different operating mode relative to the traditional structure of large base load generators (e.g. huge nuclear or coal fired steam turbine stations) with a very narrow band of peak operating efficiency and relatively slow response to changes and smaller peaking stations (e.g. diesel or gas turbine stations) which are capable of fast response to the changes in demand.
One approach to help balance this dynamic is the implementation of “Time of Use” energy tariffs. This breaks the day up into periods of high and low use, and the price of a kilowatt hour of electricity varies in relation to what time block the energy is used in. The underlying principle behind this is that in a traditional, non-solar, network there is a significant overcapacity through the night while most people are sleeping if the system was sized to accommodate the peak daytime demand. The idea is not new – when I was living in Germany in the early 1990’s this was already an established structure, and most large household appliances which did not need to run continuously (dishwashers, laundry machines, …) incorporated timers so they could be set to run overnight when electricity was less expensive, and electric water heaters had programmable cycles to optimize on energy price. The coin-op laundry in my apartment building charged different rates depending on the time of day, leading to the busiest time being 3 AM on Sunday morning because it cost about half the price to wash and dry a load at that time compared to washing only at a more normal time.
Unfortunately, time of use charges combined with a solar supply cycle on the grid don’t necessarily work as they do in a non-solar grid. To make most effective use of the solar supply, energy prices would need to peak at night when the solar supply is offline and then drop in the middle of the day when supply was highest. Given the relative percentages of solar vs non-solar supply, the overlapping model becomes quite a complicated system, particularly when one looks at the regulatory environment that this occurs in.
Here is where the Smartgrid concept comes into play. By using a network of interconnected sensors to actively monitor the systems involved, the overall grid can be continuously optimized and near-term predications made based on real-time data. As the technology grows and becomes more prevalent, the granularity of the involved systems improves – where once this was used to help determine when to bring peaking stations online, now it can be used both to do that as well as to throttle back demand by reducing discretionary loads at the end user level. This can take many forms, such as utilities providing discounts to residential customers who allow the utility to install remote disconnects for high load devices such as air conditioners or swimming pool pumps or commercial customers who enable the utility to manage their building thermostats within a given range.
This all applies to production and usage, but there is another large area still to be integrated: storage. Various forms of energy storage have been available for decades at both large and small scales, however they were generally quite expensive, limited in scope, and relatively inefficient. One example of a large scale system is a hydroelectric peaking plant: when electric demand is low excess production is used to drive pumps that send water uphill into a reservoir, but when extra power is needed to cover peak loads the water is routed back down through turbines. Variations of this have been demonstrated using air in abandoned salt mines and with giant mechanical flywheels turning in a vacuum chamber. At a smaller level some commercial users have either battery or capacitive storage capability, mainly intended as an independent system backup rather than a grid level storage, but the end effect is that depending on how these systems ae set to charge it can have a positive effect on stabilizing the grid.
What has been missing from the US grid is a homeowner level grid storage capability of sufficient capacity to aid in stabilizing the grid as a whole. Again, this technology has been on the market in other regions for years, but the US utilities have until recently been strongly opposed to the concept. What is currently happening, though, is that as homeowner scale installations of solar and / or wind energy increases, utilities in certain areas such as Southern California are finally coming around to see homeowner generators as a potential source of the overall solution rather than a problem to be excluded from consideration. By offering suitable incentives to bring the homeowner’s cost of the necessary infrastructure down to a reasonable level, the utilities hope to create enough capacity to be an effective overall tool in better managing the grid.
Which brings me back to the postcard I received. The sender turned out to be a consulting firm that represents several of the “home battery” manufacturers and performs system design activities on behalf of independent system installation contractors. After doing a bit of research to get an idea of what was available and typical pricing, I called the number on the postcard and spent a bit over an hour having a detailed technical discussion with a system designer, and we ended up deciding that a Tesla PowerWall 2 would be the most cost effective solution for my installation within the incentive program. When I initially purchased my solar array I selected a SolarEdge inverter in order to be upgrade ready for when they came out with the StorEdeg system in the US market, and while there was a LG battery which potentially could have worked with the StorEdge DC-DC coupled approach, by the time the cost of upgrading my inverter and the battery were combined the Tesla with the built in inverter made more sense even with the added efficiency hit of having to go DC-AC-DC. In addition, although currently not set up to do so, the DC-AC-DC option provides an ability to eventually use grid charging capabilities at a future point.
So what does adding a battery to my system bring me? At the moment I have to admit not a whole lot. My current system is pretty close to a net zero system, e.g. I produce basically what I use, so I don’t have much in the way of excess capacity to store. My local grid has been very reliable, and there are relatively few cloudy days or even periods within a day due to my high desert location. If anything, I actually stand to lose from both a cost and production basis, as despite the incentives the battery system will still run me about $5000 installed and whatever electricity goes through the battery is subject to around a 10% loss due to storage and transformer losses.. and I’ll still be paying the monthly minimum utility fee for being grid tied. In the event that I was living in one of the backward utility regions where they use a producer / consumer accounting system (whereby any excess energy production to your own needs is “sold” to the utility as a low rate and then any additional consumption is “bought” from the utility at a higher rate, even if the net energy comes out to be zero) instead of net metering, the situation would be different as there is a direct cost advantage to storing and using your own production.
So what do I stand to gain? In large part it’s simply the feeling of doing what’s right for the greater good. More expansively, it’s taking one of the steps within my power to actively demonstrate my opposition to Mr. Trump’s misguided and disastrous energy and environmental policies. More concretely, I do gain slightly stronger self-generation capabilities in the event of a grid outage. Without a battery system, a grid tie solar installation is required to be turned off whenever there is a grid outage in order to prevent current flowing from the solar array back through the grid to the point where it is being worked on and potentially injuring a utility worker or others who may come in contact. With a battery system part of the control circuitry will isolate the grid side as before, but will still allow the panels and battery to supply the household needs. In this case the panels will only be shut down in the event that the battery is fully charged and there is no household draw.
How does this help the utility? Quite simply, it helps them to shave off the peaks and troughs of solar production going into the grid. If every solar installation had a battery and was properly configured, the excess energy produced during the day would be stored locally and then available for overnight use. This helps to shave the peaks and troughs of solar supply and night demand at an individual site, and by extension to the rest of the grid. When these individual storage capabilities are then integrated into the Smartgrid as part of a distributed generation and storage network, it provides a dynamic energy reserve for the system to be able to draw on in. Likewise, in the event that there is a grid outage, even a few residential batteries in the affected region could be used to help buffer the step load to the remainder of the local grid as power is restored to the affected region.