The Death of a Power Plant
Replacing aging infrastructure is no joke.
One of the big pluses of living and working in a wealthy society is that power and heat are perpetually available to paying customers in the residential, commercial, industrial, and educational sectors. The university I am privileged to work for is a small city with a daytime population of about 65,000 individuals. While it is easy to think that all we need is lights on, so you can see the math on the whiteboards, and outlets to charge our laptops, the teaching and research enterprise consumes a lot of energy. We use steam to keep buildings warm in the winter, while electricity powers everything with a wire. Further, electricity demand on a university campus is strongly correlated with Nvidia’s stock price, as researchers increasingly turn to using graphics processing units, which use massive amounts of electricity and generate waste heat – driving up cooling requirements.
Meet UC Berkeley’s aging Cogeneration Plant, which we proudly operate and provides approximately 90% of the electricity and 100% of the steam needs of the main campus. Like the author of this blog post, it is aging rapidly and will need to be replaced or retired within the next five or so years. Also, it is powered by a carbon and local-pollutant generating fossil fuel – natural gas. Imagine my true excitement when I got a letter from my boss asking me to serve on a committee to help think about what we should do as a campus to procure energy services post co-gen. This excitement was quickly replaced by sheer terror, as all of a sudden everything I talked about in my classes became very tangible and real. I am writing this blog post as a self help guide to clarify my thinking, and it obviously does not represent the committee’s or university’s official position. But this is an energy economics blog. And Severin told me this was my week. And you do what Severin tells you, because we owe him.
Kerfuffle #1: Upon first thought it might be cheapest to just replace the current cogen plant with a new one. You might think the capital investment is probably lowest, and you can use the existing campus infrastructure. Plug and play! Well, if you have ever walked across our beautiful campus, you see steam leaking out of, well, everywhere. This steam is wasted energy directly and indirectly adding to global warming. Fixing the network of super old steam pipes on campus is also super expensive and not fun to fix for anyone. Second, continuing to operate a fossil fuel powered generator requires the purchase of natural gas, as well as permits under California’s cap and trade system. So regulation provides incentives to not do that. Which is why we have the regulation. High fives.
Kerfuffle #2: Political pressures, both formal and informal, push a public institution of higher learning to walk its talk and generate its energy services in as climate-friendly a way as it can. Members of our community are very vocal about the ethical and moral requirement to stop emitting carbon from our power plant. And it is not just community members – we are also required by the UC systemwide sustainability rules to decarbonize. So, no fossil plant for you.
Kerfuffle #3: The economics of evaluating different alternatives are tricky. For example, discount rates are easy to apply in theory, but trying to figure out what discount rates to use for the private portions of an investment that is going to outlive most of us and the right one for the social benefits/damages it generates is not straightforward, especially when (unlike the federal government) there are no “set guidelines” for this. The choice of discount rate in our setting is likely going to drive which technology wins out in the end. I was also super excited to run into the latest work on the Social Cost of Carbon and the Social Cost of Methane in an actual project setting – not just when I am thinking big thoughts. Thank you federal government for these numbers (and the excellent prose explaining them).
Kerfuffle #4: Reliability is critical in a university setting. You might think that an outage just means that a bunch of students might not be able to charge their phones, but there are real safety and scientific consequences to outages. Lights going out in heavily populated areas is clearly a safety hazard. Labs on campus, as we learned during the pre-pandemic public safety power shutoffs, require electricity to conduct “science”. The loss of refrigeration can lead to irreversible loss of scientific materials, like ice cores and cell cultures stored in fridges and freezers. Further critters of all sizes need to be kept comfortable, so air conditioning is key. For coding nerds like me, we use computers calculating things that take a long time to run, and if they stop, you often have to start from scratch. So having power is key and hence having control over when and where the power stays on and goes out is key. This is easier to do if you have a power plant (possibly augmented with battery storage) onsite than offsite and the ability to direct energy to the most important needs.
Kerfuffle #5: A university is a research and teaching operation. We build things and learn by doing so. A significant share of our graduate and undergraduate students are interested in doing their part to push along the energy revolution. Whatever project we finally decide on pursuing will have, in part, been informed by research done in our own labs and spreadsheets. Going forward, a new energy system on campus should provide ample learning opportunities for our current and future students.
Kerfuffle #6: Paying for such a major piece of infrastructure can be tricky. Energy infrastructure is not cheap. The state has already apportioned significant funding to help us start the first phase of the project (thank you Sacramento and people of California!). So, how to pay for the replacement infrastructure is of course on everyone’s mind. But if we go with a carbon neutral piece of technology, like giant heat pumps, we will need more electricity than we can generate locally. So we will buy (low/no carbon) electricity off the grid, which we have to pay for to operate our gadgets. Hence we need to be thinking about energy price risk – in a finance sort of way. In order to make good decisions, we need to have a good understanding of the future distribution of energy prices and prices of cap and trade permits in California. Making the best decision from an economic point of view is going to involve not only a best guess of “one number”, but a characterization of the range of possible price outcomes.
Kerfuffle #7: Incentives clearly matter. In a world of growing energy demand, setting the right incentives for energy users on campus is key. Our campus is gorgeous, but our buildings are not new. Severin, Meredith and the Jims have thought a lot about optimal energy pricing, but this assumes good measurement. How do you do this when you can’t meter individual users in an environment where big buildings have very heterogeneous user populations and consumption patterns?
So while this is not my funniest or most interesting blog post, I am really enjoying being part of this task force, since it makes what we do here so very real. Addressing these kerfuffles will not be easy, but it also makes me realize how far we have come in terms of providing real numbers used in day-to-day decision-making in institutions large and small. High fives fellow energy and environmental economists. I am proud of us. Let’s keep up the good work.
Follow us on Bluesky and LinkedIn, as well as subscribe to our email list to keep up with future content and announcements.
Suggested citation: Auffhammer, Maximilian. “The Death of a Power Plant” Energy Institute Blog, UC Berkeley, February 26, 2024, https://energyathaas.wordpress.com/2024/02/26/the-death-of-a-power-plant/
Categories
Maximilian Auffhammer View All
Maximilian Auffhammer is the George Pardee Professor of International Sustainable Development at the University of California Berkeley. His fields of expertise are environmental and energy economics, with a specific focus on the impacts and regulation of climate change and air pollution.
Energy Institute Blog 
Steam distribution systems are incredibly inefficient. Arguments are made that they are efficient because they employ cogeneration technology and use waste heat. However, the losses associated with steam distribution systems are very high. Many years ago, I worked at a utility operating a steam utility and we estimated the losses as exceeding 50 percent. In contrast, hot water distribution systems are much more efficient but would require replacing the distribution system as well as retrofitting the buildings to accept hot water instead of steam. Given the limited heating load in Berkley (I am a northern Midwesterner), I suspect a centralized heating system no longer makes sense.
Hi Maximilian
Looks like quite a project, should be very “interesting”. Congratulations.
My two initial reactions regarding scope and timing:
Scope> AS I think you know already, the task at hand is not just about replacing the power plant, but the creation of a long-term strategy for serving the campus’s total energy needs. I suggest you frame it that way in the future so you can best communicate the enormity of the project.
Timing> This likely a 10-year project, not a 5-year project. iI you let yourself be bound by a 5-year clock it will be hard to evaluate your options on a fair basis, make decisions, and implement them. With that in mind I hope you can find a way to extend the life of your co-gen plant a few years.
I have to wonder, doesn’t the campus already have an energy master plan that addresses many of the issues raised by the commenters? I’d be surprised if it did not.
Regards.
Across the street (on UC land) there is a DOE lab running a top500 supercomputer and a particle accelerator (synchrotron) among many other energy intensive labs; They get power from the grid, and they don’t pay regular PGE prices; how about running a wire from there?
And how about using all these GPUs to warm up the swimming pools? (liquid cooling is a thing now).
+10 for denoting some of the Kerfuffle’s.
A few trips over to Standford to see how their plans/solution(s)- for the replacement of their co-gen central power station- have played out may be in order.
Many campuses are replacing old Cogens with new age Thermal cooling and heating networks (plus 100% clean power procured from renewables + plan for onsite reliability backup generation for critical loads). My collogues and I coined the Thermal Microgrid term for it.
https://www.edf-innovation-lab.com/wp-content/uploads/2018_SESICaseStudy_FINAL.pdf
https://www.edf-innovation-lab.com/thermal-microgrid-project/
As you are probably aware, UCB’s daughter campus UC Davis has already headed down this road on replacing its steam system. And it’s supposedly reducing fossil fuel use at its new Aggie Square campus (although I’ve heard there may be deviations.) Here’s more on this: https://www.ucdavis.edu/climate/news/uc-davis-finalizes-fossil-fuel-free-pathway-plan
I worked at Berkeley Campus in the 90s and the Cogen plant was under my management. Our team re-wrote the Cogen Contract, rewired the campus as well as conducted repairs on the steam heating system. At that time, the 25 MW Cogen just met the electrical needs of the campus. The PG&E electricity grid hook up was also sized to meet that load. Of all your Kerfuffle comments, Number 6 with the statement: “buy (low and no carbon) electricity off the grid” appears to be the most risky and expensive variable. The Berkeley Campus’ electrical load may now be approaching 50 MW in the next 10 years. I note that the current plan for that campus is “An additional 15 MW of on-campus solar power, 8 MW of fuel cell power, and 30 MW/h of battery storage capacity will provide base power and resiliency to the system.” This makes the reliance on the PG&E grid’s low carbon power to be significant.
Gary Matteson
Some parts of the SF Bay Area can exploit (sub)urban solar parking lot canopies located at ubiquitous & widely distributed large apartments, condos, shopping centers, business parks & various public facilities, but that doesn’t look like a significant NetZero opportunity for UC Berkeley.
Anonymous might be right, if rehabbed cogen can work for another 20 years. Beyond that time frame, the entire SF Bay Area needs to tap into the massive north coast offshore wind resource & develop some new, strategically distributed 24/7 Advanced Geothermal power plants. Don’t laugh, it’s happening, but it’s going to be an expensive & time consuming process. An on-campus Advanced Geothermal power plant at the Lawrence Berkeley National Lab would be great…….maybe in 2045? Talk to Secretary Granholm. She’s really high on AGS.