This is the last article in a four-part series about the unique O&M challenges of the energy storage asset class and the reasons why owners and operators of mixed portfolios of solar, wind and energy storage assets need an integrated asset performance management platform.
In this article series so far, we’ve been discussing how the battery storage asset class introduces financial risk to hybrid renewable power projects due to uncertainty about the true cost of running this asset over the life of the project.
In this article, we will talk about how an integrated data platform can be used to support the evaluation of financial tradeoffs of hybrid renewable power systems when energy arbitrage is added to the equation.
Adding an energy storage system into a solar or wind power facility introduces new project complexity. First, the plant is more technically complex: I have many more operational choices to consider than I had before. Second, it introduces additional financial complexity: I have many more buying and selling decisions to make than I have with a stand-alone wind or solar plant.
Let’s look at each type of complexity that the asset manager and scheduling coordinator needs to consider when operating this new asset class in conjunction with a wind or solar plant and review the kind of decision-support information each of these roles need to perform their job.
The operation of a stand-alone solar power plant is pretty straightforward. When the sun is shining, our goal is to make as much power as possible and export it to the energy off-taker. When the sun isn’t shining, the plant sits idle and perhaps we perform maintenance on it to minimize production losses.
The same cannot be said when we add an energy storage asset to the mix. It may be operationally advantageous to charge the battery system during the off-, mid- or even on-peak energy price periods during the day – or it may not.
Why? Every change/discharge cycle takes a toll on the battery’s life and the frequency and depth of discharge of the battery can dramatically affect its useful life. Therefore, I first need a good understanding of the cost of charge relative to the revenue gained from discharge.
In addition, the battery supplier warrants the equipment only if the operator stays within a specified operating envelope. Exceed those operating limit conditions and the warranty may be invalidated. If there is only so much energy that can be processed through a battery and only so many cycles over the course of its life, how can I as an asset manager know if I am optimally consuming the battery?
To answer these questions, I need up-to-date and very detailed information around the operational performance of the battery, its rate of capacity fade and its estimated useful life. Without this information, the optimal scheduling of the battery into the market is no more than informed guesswork.
As suggested above, every operational decision for a battery storage device is ultimately a financial decision. That’s because a battery storage system, unlike its renewable power cousins, the wind and solar asset, is consumed, not run. There are only so many cycles in the life of a battery.
The key question is: How are we going to use that capacity and cycles? Each time we charge or discharge the system we are reducing the life and future use of the system. Therefore, every time the battery is cycled, we need to consciously ask, “Is this the best next financial use of the system?”
Depending on the off-take agreement of the facility, this decision may be made for you by the utility or you may be able to make the decision on your own. Regardless, the battery has one fewer cycle and one unit of reduced life in it. Does the revenue from the cycle exceed the cost of storage? If I defer this cycle will there be positive incremental profit in that deferred cycled? Does the overall project net present value (NPV) go up or down if I cycle now – or should I defer based on my best estimate of future energy price strips? These questions and many others need to be asked by an asset manager every time the system is dispatched into the market.
Ultimately, the decision to charge or discharge a battery storage system should be based on the results of scenarios run through an economic optimization model. Whether the optimization evaluation uses advanced analytics data and a formal what-if tool or a spreadsheet analysis, the asset manager is running an economic optimization routine to guide that decision.
When evaluating the economic optimization of an energy system, I like to consciously consider the “ABCs” of optimization:
The “A” in the ABCs of optimization asks the question, “What can I adjust?” Optimization specialists call this the problem’s “degrees of freedom,” or decision variables. For a battery storage system, these are the operational and commercial input assumptions into the what-if model.
For example, how often I cycle, how deeply I discharge the battery and whether I sell or store power are just a few of the operational and commercial “knobs” I can turn to evaluate future battery scheduling scenarios. The asset manager and scheduling coordinator need to be armed with the best and most current information about the battery’s operational performance and cost structure to properly characterize what can be adjusted for future dispatch scenarios.
Another question we need to ask is, “What is best?” Usually, when solving economic optimization problems, what’s best is profit maximization. Depending on the off-take agreement, the best solution might solve for maximizing revenue or minimizing costs. However, usually the “objective function” in optimization is: Which dispatch scenario returns the greatest profit?
The next thing to consider in optimizing an energy system is characterizing the system’s “constraints.” Constraints are both technical and commercial. For example, the technical limits imposed on the equipment by the warranty establish some of the physical constraints of the system. In accordance with the warranty, the battery can only be cycled so many times per day or year, the battery C-rate must be maintained, and its state-of-charge must stay within a certain range. These operational constraints define the physical limit of charging and discharging. It doesn’t matter if real-time energy prices just went to $500/kWh; if you have already reached the daily battery cycle limit, one more cycle will violate the warranty.
Battery system commercial limits are usually defined by the power purchase or interconnection agreement. If the battery is selling into a capacity market and has a certain rating, it doesn’t matter that the battery has additional capacity available, the seller won’t get paid for selling more capacity than contracted. This is a commercial, not technical, constraint. Optimization constraints establish the multi-dimensional walls that the system can play in — no going outside the walls.
It should be clear from the discussion above that solving the lifetime economic optimization problem of a hybrid renewable plant is not trivial, especially if the battery system can participate in a real-time energy market. Whether the asset manager is evaluating the future dispatch scheduling problem with a spreadsheet or a sophisticated what-if model, the model needs to be constantly fed with up-to-date operational performance and cost data if it is to properly evaluate the financial tradeoffs that result in the best dispatch schedule.
With this and the previous articles in this series, I’ve made the case for why a fully integrated asset performance management (APM) platform is essential for feeding the project’s what-if model with the data it needs to derive the optimal dispatch schedule for a battery storage system. The asset manager will need to consider how often the model’s input assumptions and performance model needs to be updated and whether model assumptions should be manually updated or automatically updated by the APM. Regardless, without regular data feeds from the APM into the economic model, economic optimization of a hybrid renewable power project that participates in a real-time market is not possible.
The introduction of the battery storage asset class brings increased complexity to the renewable power asset owner. To properly dispatch this new asset class into the market requires that many economic tradeoffs be evaluated. Lacking integration of the operational performance, operating cost and life estimate data, asset managers and scheduling coordinators are ill-equipped to schedule the storage system successfully into the market.
Drive O&M and Drive APM work together to empower teams with a condition-based maintenance solution purpose-built for renewables.
Interested in learning more? Download the Drive O&M data sheet.
Now retired, Steve Hanawalt co-founded Power Factors in 2013 and spent almost 40 years in the energy industry. For more renewables industry content, follow Power Factors on LinkedIn and Twitter.