Variable Generation & Operational Complexity - An Illustrative Example

Wind and solar generation are providing increasing volumes of energy on power systems around the world as capital costs for these technologies decrease and carbon prices and government policy support increases. This is great for reducing the overall carbon emissions of power systems, but there are operational implications that must be recognized, planned for, and managed. In this article, I'll use some actual data to demonstrate the variability these resources introduce and the challenges this creates for power system operators.

My intent here is not to disparage or criticize variable generation. Renewable resources are variable by nature, this is not a value judgement, it is simply a statement of fact. And it is a fact that we must plan for and learn to manage. My intent with this article is to provide an illustration of the operational challenges these resources introduce to help create awareness for the need to plan carefully and holistically for the impacts of these kinds of resources on our power systems.

The Balance Between Supply and Demand

A unique characteristic of power systems, and one that creates significant operational complexity, is that the supply of power must be very precisely balanced with the demand for power instantaneously. For most of the 100+ year history of alternating current (AC) power systems, (the operation of the Niagara Falls generating station with long-distance transmission was the first "grid" in 1896) the largest system variable was demand because the generators, almost exclusively synchronous, could be controlled to relatively tight limits but when a power user turned on a light switch or a motor could not be precisely predicted or controlled. This meant that system operators mainly had to worry about adjusting generation output to match instantaneous variation in the demand for power. However, in today's power systems, an increasing proportion of power generation comes from renewable resources with variable output that cannot be controlled by its operators. When added to the inherent variability of power demand, the increasing variability of power supply adds significant complexity and operational burden to power system operators. This must be acknowledged and planned for as we evolve our power systems toward net zero.

Real World Example of Variability

An excellent example of the operational challenges of supply variability is the relatively new Travers Solar Project in Vulcan County, Alberta. For the owners of Travers, please understand that I am not singling out this facility as a criticism of the project, I am focusing on Travers because it is the largest solar generating facility on the Alberta power system and thus illustrates the impacts of variability at scale and makes them more obvious than with smaller scale facilities.

This 465 MW solar facility is at the limit of Alberta's Most Severe Single Contingency (MSSC). The MSSC limit for the Alberta power system is 466 MW when Alberta is connected to BC through the AB-BC intertie and 425 MW when Alberta is disconnected from BC (islanded) which occurs frequently. The AESO is working towards maintaining the 466 MW MSSC under all conditions through their Reliability Requirements Roadmap and Market Pathways initiatives, but this is not relevant for the illustrative purposes of this discussion. Suffice it to say that the Travers Solar Project is a very large single source of power on the Alberta power system and changes in its output have significant operational impacts.

Observing the output of the Travers facility reveals a large amount of variability, which is, of course, determined by the weather conditions. Here is a sample from October 11, 2023 using screenshots from the app Dispatcho:

On October 11th, solar output was rising up until around 13:00 and then a period of extreme output variability occurred. There were 6 large ramps between about 100 and 70 MW over the span of an hour and 50 minutes followed by a significant ramp up from 76 MW to 240 MW in 17 minutes (9.6 MW/min) followed by a ramp down from 240 MW to 60 MW in 22 minutes (8.2 MW/min). For context, keep in mind that this is just one facility and does not take into account the variability in other solar and wind facilities on the power system during the same span of time.

Now consider the perspective of the people responsible for managing the power system, the System Controllers (SCs) in the control room for the Alberta Electric System Operator. These are the unseen unsung heroes of the grid, the people that literally "keep the lights on". The SCs must balance supply and demand in real time while also ensuring the system's frequency and voltage remain stable. They are used managing system variability but the kind of variability from Travers on Oct 11th is far beyond what would be considered historically "normal" for a large generator. Beyond the natural response of the power system from the online synchronous generators providing inertia to counteract changes in power flows, the main tools the SCs have at their disposal to maintain system reliability are called Operating Reserves - regulating reserves, spinning reserves, and supplemental reserves - in addition to more localized actions to manage voltage. A simple way to think about operating reserves is that they represent unloaded "headroom" on generators that are either online or offline that can be called on by the SCs to compensate for unexpected supply-demand imbalances in a faster timeframe that re-dispatching generators through the real time energy market. On Oct 11th, the kind of up-down variability demonstrated by Travers could only be managed through regulating reserves as these respond automatically and can help manage variability in both the positive and negative directions and the SCs likely would not have been able to direct spinning reserves on and off quickly enough to manage these output variations. However, while regulating reserves may have automatically helped the SC's manage this situation, they also may have had to take local actions with other generators and utilities in the Travers area to manage voltage issues. Suffice it to say, I can only imagine that Oct 11th was a very challenging day for the AESO SCs!

Implications

"So what", you may be saying, "isn't this exactly why we have operating reserves?" Well yes, but remember that this was just one solar facility on one day. And given the fact that Travers is at the MSSC for Alberta, this means that a portion of Operating Reserves will likely have to be procured specifically to manage the large variability of this one facility. Now consider that there are many existing and many more future solar and wind facilities in southern Alberta and all of them are exposed to the same wind speed and solar intensity changes. On aggregate, this will mean a lot more variability on the system. This will mean the AESO will have to procure more Operating Reserves, which are very expensive and will add to the cost of system operations, and it will mean we will need new ancillary services, such as Fast Frequency Response, to manage sudden losses of power supply from large variable generators and the AB-BC intertie. And this doesn't address voltage control, which requires localized actions and responses. All of this will further increase operational complexity and the cost of operating the power system. All of this will be paid for by power consumers - you and me.

Even More Complexity

Now that we are seeing more and larger solar facilities on the system, we are learning about other nuances that increase complexity. For example, many solar faculties, like Travers, have articulating solar panels. Solar facilities use articulating solar panels to ensure maximum solar exposure for each panel to maximize the power output from the facility. This makes perfect sense from a design perspective and is what any good engineer would do - maximize output.

The ability to articulate the panels provides an additional benefit. When there are high winds hitting solar panels that are mounted to the ground, the force of the wind puts significant stress on the panels that can tear them from their bases if the wind speed is high enough. Solar facilities with articulating panels can use this capability to reposition the panels to reduce wind stress and hence protect the panels from wind damage. This is a great benefit for the longevity of the facility, but it has an unintentional impact on power system operations and it is a great example of the unintended outcomes that can occur in complex systems. I suspect no one considered the impacts to power system operations because facility design criteria are specific to a facility and typically would not consider broader system effects.

Southern Alberta is blessed with good wind and solar resources and this is why we are seeing concentrated development of wind and solar generating resources in this area. What this means for power system operations is that all of these variable resources will be exposed to weather variations at the same times. This has both negative and positive implications for system operations. The negative implication in that SCs can expect large decreases and increases in all solar generation when clouds blow through the area or large increases and decreases in wind generation when weather fronts move through the area that increase or decrease wind speeds. The positive implication is that, over time, we can generally understand the characteristics of the weather patterns in this area and, to some degree, begin to approximately predict and prepare for the impacts of weather variability on solar and wind resources.

Now, introduce articulating solar panels to the picture. When a sunny day coincides with high wind speeds from the right direction and variable cloud cover rolling through the region, this could result in a very difficult to anticipate additional source of solar generation variability. If a solar operator has to make a series of solar panel position adjustments to protect the facility from high winds, and if there are clouds passing through that vary solar output, articulating the panels away from wind will decrease solar exposure and decrease power output, which adds to an a very unpredictable output pattern from the facility. No imagine if this is happening at multiple facilities across the system. And because this will depend on wind speed, wind direction, and cloud cover, it would be very difficult for the SCs to predict and prepare for.

Now this may be manageable if it was just one facility, but as solar generation design becomes more sophisticated, we will likely see more optimization like this and the compound effect on the power system is even more operational complexity and challenge. And articulating panels is just one design nuance - I imagine there will be more nuances we haven't considered yet.

Conclusion

Now, I don't know what the specific factors were that drove the extreme variability observed on Oct 11th at Travers. All I can do is observe the publicly available data to understand the impacts to power system operations. However, I point to this example as a real-world-right-now illustration of the fact that variable generation sources introduce many factors, both obvious and not so obvious, that increase operational complexity.

As we see more and larger variable generation faculties on the system, we must dive much deeper into the design and operational characteristics of these faculties to ensure we identify system operational challenges and plan for them BEFORE they connect to the power system. This will help ensure connection standards and requirements are sufficient, operational models are accurate, and operating reserve procurement is sufficient to maintain reliability and security. We need to understand how large scale storage can help both at the facility level and at the system level. And we need to makes sure we minimize the cost of system operations that will ultimately be paid for by customers.

From my perspective, the operational complexity that variable generation introduces is further support for the need to move to an integrated resource planning (IRP) approach for the Alberta power system. We need to ensure we have the right transmission and generation resources in the right places, with the right sizes and characteristics, and at the right times to ensure ongoing reliability and operability as we evolve towards net zero. A planning process that optimizes between transmission and generation will accomplish this. To read more on my perspective on IRP, please see my article on the topic.

Hopefully this was helpful in illustrating the increasing complexity of power system operations we are facing in the drive to net zero.