Assembly Standing Committee on Privacy and Consumer Protection

Good afternoon.

Good afternoon. I am calling to order this joint hearing of the Assembly Committee on Utilities and Energy and the Assembly Committee on Privacy and Consumer Protection. We we are here today for an informational hearing to discuss the energy impacts of AI. Before we begin, I have just a couple of housekeeping announcements.

First, as is customary, we will not accept disruptive behavior and will apply Assembly rules to maintain order and run a fair hearing. Second, public comment is welcome here in person and otherwise it may be submitted online via the Committee website. So with that, let's turn to the topic at hand.

The surge of investment into AI data centers has been noted as bigger than the build out of the interstate highway or the.com boom. As we all know, electricity availability and affordability has been a top concern shaping this data center development.

The scale, speed and uncertainty of this new load are unlike prior waves of electrification, but the potential to challenge long standing assumptions embedded in energy forecasting, planning, cost allocation and rate design. The core question, to be clear, is not whether data centers should be served, but under what terms?

How do we fairly design rate structure and how do we ensure that California consumers are protected from any potential risks? Thus far, California has largely avoided rate impacts from data center growth, but that outcome is certainly not guaranteed.

Colleagues, I believe that the task before us is to ensure that California is positioned to welcome this innovation without repeating mistakes that we say that we see playing out in other regions of the country. We have a rare and important opportunity to reassess existing rate design and customer protections.

The promise of data centers is alluring that increased electricity sales from these data centers will cover all the new costs to serve both those data centers as well as offset existing grid costs. And I do believe that if managed efficiently and effectively, this enormous growth can indeed be beneficial.

Today's panelists include a range of experts from researchers to utilities to agency professionals and to the data centers themselves. They've been asked to speak to needed safeguards to ensure that data center development in California does not lead to runaway energy costs, stranded assets or negative environmental impacts.

I'm looking forward to hearing from today's panelists as well as from the public. With that, I will go ahead and turn to my co Chair, Assemblymember Bauer-Kahan, the Chair of the Assembly Committee on Privacy and Consumer Protection.

Well, thank you, Madam Chair, and thank you for putting this together. You know, I want to say to you, Madam Chair, that you have been a true force for protecting California's ratepayers and ensuring that our grid that we are building out for the clean energy Future of California is sustainable and reliable.

And I think this is an important part of the conversation. And the reason the Privacy and Consumer Protection Committee is here today is our job is to ensure that California's innovation economy continues to thrive while also protecting Californians in that goal.

And so, as you heard from the chair, that is really the work that we have before us today. Nobody thinks that we should have a future without data center growth. It is important part of our technology economy, but we need to do so in a way that doesn't hurt Californians.

I want to start by thanking your staff, the Utility Energy Committee staff for their hard work on this and of course the staff of the Privacy and Consumer Protection Committee for their partnership Rules Committee, the sergeants and everyone else that puts these hearings together. Across California, energy intensive data centers are being built to support the AI industry.

As we've mentioned, and according to data center map, California has over 280 data centers, making it one of the nation's leading states for data center development, second only to Texas and Virginia. The data centers House critical computing infrastructure. We'll hear today from my very own Lawrence Livermore Lab.

There couldn't be infrastructure more important to our national security than the lab itself and the data center it houses, which allows for high intensity computing and extensive data storage for AI training.

But it increases energy demand, as you have heard, and requires us to expand our grid in ways that benefit both consumers, but also these larger load users. Data centers and their impact on the affordability of electricity have become a political flashpoint across the country.

But it's important to note that we are seeing bipartisan support that these data centers are both important and need to have regulations around them to protect Americans and the energy economy.

In California in 2024 alone, data centers added an estimated 50 gigawatts of new capacity to global development pipeline, which is roughly the entire electricity demand of the state of California during our most extreme heat waves. Despite being a top market for data centers, we have not yet seen an increase in electricity rates according to our cpuc.

So that means this is the perfect time to be having this conversation because we can plan ahead and make sure that we don't harm California's ratepayers while continuing to move forward with the data centers, the innovation economy they support and ensuring grid reliability.

As many of you know, I've worked on the issue of transparency around these data centers and that is in part because I think it is critical that we build a grid to keep the lights on for everybody. And that includes the data centers who need the energy to be consistent as well.

And we can do so in a way that is responsible and clean. So I am excited for the conversation today. I want to thank all of our panelists for being here. And I'll turn it back over to Chair Petrie Norris.

Thank you, Madam Chair. All right, so to begin today's hearing, I would like to welcome Nate Gleason, who is the program leader of Cyber and Infrastructure Resilience at Lawrence Livermore National Laboratory. And we've asked Mr. Gleason to join us today, really to frame the discussion before we jump into our panels. Welcome and thank you so much for being here.

Thank you. Thank you, Chairs and Members of the Committee for the opportunity to testify today once again. My name is Dr. Nate Gleason. I'm the program leader for the Cyber and Infrastructure Resilience program at Lawrence Livermore National Laboratory. We're a Department of Energy National Laboratory, home of the National Ignition Facility and the world's largest supercomputer.

We know the importance of the state having reliable, sustainable and affordable energy, given California's leadership in national innovation and the lab's long standing presence in the state. For decades, utilities have planned for roughly 1% annual growth in electricity demand. Efficiency improvements have made even this an overestimate.

Essentially flat demand has resulted in deferred transmission investment and generation retirements without full replacement. That era is ending. The Department of Energy estimates that we need an additional 100 gigawatts of peak supply in the next five years across the United States, with about half that demand coming from data centers.

Data centers represented 4.4% of total US electricity consumption in 2024. By 2028, that could grow to between 6.7 and 12% in California alone.

The CEC has reported requests from data center developers for 18.7 gigawatts of new, which is almost 40% of the current peak demand, and has recently predicted that electricity consumption could rise by more than 50% by 2045.

Data centers represent the physical backbone of an artificial intelligence revolution that, with initiatives like the Department of Energy's Genesis mission, are poised to drive breakthroughs in science, energy, and national security. Breakthroughs that can happen here in California. There are real strategic advantages to hosting data centers domestically, keeping sensitive data out of adversarial hands, maintaining an economic competitiveness.

However, data centers create fundamental planning and operational challenges that differ from traditional load growth and affect grid planning, reliability, and affordability. There's three key challenges with data centers. First of these challenges is planning uncertainty. Data centers can be built in 18 months and consume hundreds of megawatts, the equivalent of over half a million households.

It can take up to 10 years to build the transmission and generation to support them. So system planners must make billion dollar investment decisions based on highly uncertain demand projections. At a macro level, there are no reliable models for predicting data center energy demand and the estimates vary widely.

There's near term uncertainty in how many data centers will actually get built and longer term uncertainty in what their workloads will be for individual facilities. System planners may not have information on attributes like behind the meter generation capacity, operational flexibility and demand response capabilities. The second challenge is around reliability.

Data centers differ fundamentally from traditional loads in two ways. The speed at which consumption can change and the lack of correlation with observable grid conditions. Traditional loads respond primarily to weather patterns that operators can observe and predict.

Data centers can change consumption in minutes or seconds, and these changes are driven by factors that may be invisible to the grid operator. Things like software deployments, training cycles and algorithmic load balancing. This can result in very rapid changes in demand on the order of hundreds of megawatts or more, which can pose reliability challenges for system operators.

The third challenge is affordability. New large loads increase resource adequacy requirements and can create or worsen network congestion, forcing more expensive local generation and and driving wholesale price spikes that ultimately are passed on to ratepayers. Overestimating demand results in stranded assets whose costs ratepayers bear through higher rates. Underestimating leads to price spikes and reliability risks.

Even data centers with substantial on site generation may still draw significant power during peak periods and when transmission built for a data center provides secondary system wide benefits. Existing customers may be allocated costs for infrastructure they didn't request and wouldn't need without the new load. Addressing these challenges requires moving beyond traditional planning approaches.

Planning around any single demand projection will almost certainly be wrong given the unprecedented uncertainty in data center growth. Planning methods need to be robust across many possible futures rather than optimal for one assumed and unlikely to be realized future.

Specifically, utilizing our high performance computing and simulation capabilities, we have found that stochastic optimization or optimization under uncertainty offers a promising approach by evaluating entire ensembles of scenarios to develop plans that hedge across many potential futures. This helps avoid the waste of both overbuilding and the price spikes in reliability risks of underbuilding.

Effective planning must also include concurrent optimization of generation, storage and transmission, recognizing that those decisions are interdependent. Our research has indicated that co optimization leads to significant improvements in system performance. Equally important is studying options to deploy data centers in ways that benefit the broader grid and improve reliability for all consumers.

Operators may do things like Fund infrastructure upgrades. By adding operational flexibility. Data centers can provide demand response, curtailing demand during high congestion periods. At Lawrence Livermore, we have capabilities in these areas that are being used to support the Department of Energy power grid capacity expansion analyses, both other places in the continental United States and Puerto Rico.

We welcome the opportunity to support California as the state navigates this complex decision space. I appreciate the opportunity to speak with you today and I'm happy to discuss our research in more detail in response to your questions here today or in a separate session.

Thank you and thank you. So what we'll go ahead and do is open it up for any questions for Dr. Gleason before we move to our next panel. Questions for Members.

Thank you, Dr. Gleason. And I think it's super helpful to hear how your capability is being leveraged to better understand where our future is going so we can plan for grid that isn't wasteful or under prepared.

I know that we've had conversations with the lab around how we can better coordinate and so I really appreciate the offer for you to provide your incredible resources to us as a state to allow us to have this level of intelligence around our planning.

So I hope not only we as a Legislature will leverage that, but that the Administration, the Puc, the CEC will lean into what to me sounds like some really great technology and brilliance that can help us better plan.

And so doctor, if I understood kind of the crux of what we need to do differently, it's. You used a bunch of words that I didn't understand. That's, you know, sort of smarter planning is very important. So what would that, what does that look like in practical terms?

And are there other states and jurisdictions that have moved to. You talked about stochastic planning, co optimization. Are there other states and jurisdictions who have started to plan that way in ways that California can learn from?

I'd say so, yeah. The term stochastic optimization, which is really a bit of a mouthful, meaning trying to find an optimal solution when you're very uncertain about what may actually happen. It is a area of active research. One of the challenges in looking at problems using stochastic optimization is it's very computationally intensive.

That is one of the reasons why that's one of the things we do at Lawrence Livermore.

We have again, as I mentioned earlier, very substantial resources in high performance computing that are leveraged to be able to do the very computationally intensive challenge of that optimization, taking into account uncertainty in demand projections, taking into account uncertainty in consumer behavior, in environmental conditions, in other things, it is something that is more emerging.

I think California has an opportunity to be an early adopter in this space and really use lean into that kind of capacity expansion planning. We have used it. One of the places where we have been applying it is in Puerto Rico.

So through work with the Department of Energy, we're working to help Puerto Rico rebuild their energy system in a way that not only works today, but will work for the next 10, 20, 30, 40 years as their demand evolves, as the environmental capacity conditions around them evolve.

So that'd be one example of where it's being used right now.

Okay. And then I think the other thing that I took away from your comments, it seems to me like we also need to be more nimble and more dynamic in our planning that this new world. And correct me if I'm wrong, this new world means that we can't have a five year planning cycle.

We've got to be able to be more nimble and dynamic in order to respond to market changes that are in a market that's going to be moving much faster than previous waves of innovation.

Yes, that's correct. One more word I would add onto that is more integrated. And so we can't plan out in pieces because when you do things in pieces, you don't come up with the full solution. There are challenges with being faster with planning. As I mentioned earlier, data centers can be built in 18 months.

It can take a decade to build some of the grid resources. But there are options you can do in the interim.

There are things like flexible connection agreements that I know a number of the IOUS are working on or are offering that provide a bridge to avoid some of the challenges with connecting those large loads in the near term before the infrastructure can and get built out.

But there are other things like thinking about co locating storage with the data centers to level out peaks or provide grid services to the system. These are the types of things that will enable data centers to not be a drag on the grid, but actually be a benefit to the grid and to consumers.

Well, thank you again, Doctor for joining us. I think we're going to turn to our next panel but look forward to continuing this conversation and continuing to work with you as the conversation continues. Thank you. All right, so with that we are going to go ahead and welcome a very diverse panel.

So we are going to be joined by Alicia Gutierrez, who is the Director of the Energy Assessments Division at the California Energy Commission.

Leuwam Tesfai, joining us from the California Public Utilities Commission Neil Miller, who is joining us from the California Independent System Operator Michael Medeiros joining us from Pacific Gas and Electric Nico Prokos joining us from Silicon Valley Power, and Carib Onder joining us from the Data Center Coalition. So welcome and thank you all for joining us today.

Before we jump in, just Members to kind of frame the discussion, what we've asked our panelists to do is focus the first part of the discussion on grid impacts, and then we will move into an integrated discussion of ratepayer impacts.

So we're going to start with the discussion around grid impacts, and then many of our panelists will stay for the subsequent panel. With that, I think, Director Gutierrez, I believe we are beginning with you. Welcome. Thank you.

There we go. Is it working?

All right. I do have slides, if we can pull those up.

Thank you. We can go ahead and advance to the next slide. Thank you. Good afternoon, chairs and Members. My name is Aleecia Gutierrez. I'm the Director of the Energy Assessments Division at the California Energy Commission and it's a privilege to be here today.

As you mentioned, data centers are emerging as a growing driver of electricity demand in California. Today, I'll provide an overview of how the CEC forecasts data center demand, how that forecast supports statewide planning, and the challenges associated with planning for these large loads. Next slide.

The CEC's role with respect to data centers is primarily through our responsibility for developing the statewide electricity system forecast.

Electricity demand forecast While the CEC also has authority related to thermal thermal power plant licensing, small power plant exemptions relevant for backup power, and clean energy projects such as Upton and building efficiency standards that impact data centers, today I'll focus on our demand forecasting role. Next slide.

The CEC's energy demand forecast is used by the CPUC, CAISO and utilities for electricity system planning. And the CEC produces two main forecasts. The planning forecast supports resource adequacy and integrated resource planning, as well as system level transmission planning. And the local reliability scenario is used for local transmission and distribution system planning.

The local reliability scenario yields a higher energy demand than the planning forecast to account for uncertainty when the statewide forecast is applied to more localized planning needs. Next slide. This slide provides context for where data centers fit within the broader electricity demand forecast.

While overall system demand is influenced by many factors, data centers represent a new source of growth that warrants dedicated analysis. Currently, data centers account for about 1,000 megawatts, or 1 gigawatt, or 2% of the CAISO peak demand. Under the planning forecast, that share increases to about 9% by 2040.

And under the local reliability scenario, it increases to about 11%. Next slide. This slide shows the data center capacity. Let's see. Did we go? Did we skip one? Okay. Okay. So we're on the data center methodology. Okay, so the data center forecast is developed using a bottom up project based approach. Go. Go. One more. Sorry, I was.

There we go. First, we collect data center energization requests from utilities including application status, requested capacity, timing and location. Second, we apply a set of assumptions to translate requested capacity into expected maximum demand. These include a utilization factor, confidence levels that vary by project status, and ramping assumptions that reflect when data centers reach full load.

Finally, we apply an hourly load profile developed using interval meter data from existing data centers to estimate each project's contribution to system peak demand. Next slide. This slide shows data center capacity Requests reported by utilities as of December 2025 grouped by project status. In the CAISO, there's over 23,000 megawatts of new data center capacity requests.

About half of that total consists of early stage inquiries that have not yet submitted an application. The remainder reflects projects with active applications or signed agreements which have a higher likelihood of being built and therefore contribute more to the forecasted demand.

As shown here, the majority of requested capacity is concentrated in a small number of utility service areas, particularly PG&E and SCE. Next slide this slide shows the data center demand forecast under the planning forecast and the local reliability scenario, with results broken out by project status.

The stacked bars illustrate how requested capacity at different stages of the energization process contribute to the final demand forecast. Projects with signed agreements contribute the majority of near term demand, while projects at the active application and inquiry stages are increasingly discounted to reflect uncertainty and timing and completion.

Under the planning forecast, data center demand grows to about 4,900 megawatts by 2040, and under the local reliability scenario, demand reaches approximately 7,400 megawatts over a similar timeframe. You'll notice that the growth flattens in the mid-2030s in both cases.

This plateau reflects the decision to base the forecast on currently known projects in utility energization cues rather than assuming additional future growth beyond what is visible today. Next slide Very briefly, I've included this slide for reference.

It shows the same data center demand growth under the planning forecast and local reliability scenario, but broken out by service area rather than project status. Next slide.

And finally, I wanted to briefly highlight some of the challenges associated with forecasting data center demand. First, data center forecasting is limited by a lack of historical operating data, particularly for AI focused data centers. This increases uncertainty around growth rates and operating characteristics.

Second, the forecast relies on developing confidence levels to account for uncertainty and whether and when projects ultimately come online. Third, utility energization data changes frequently, meaning the forecast can only reflect a point in time snapshot of known projects.

Fourth, because CAISO's transmission planning process looks several years ahead, it is important to include even lower confidence projects to support proactive system planning. And finally, the forecast has limited visibility into longer term growth in future data centers, especially those with AI centric workloads and those may operate differently than those of today as technologies evolve.

Next slide in closing, the rapid growth of data centers presents new challenges for demand forecasting and grid planning. By continually refining our methods and incorporating new data, the CEC aims to provide transparency and support proactive, reliable grid planning thank you. And I'm happy to answer any questions.

Thank you. And I think, Members, what we'll do is hold questions once we've heard from all of our panelists. All right, next up, we are joined by Luan Tesfai from the CPUC.

All right, so good afternoon, Chair Petrie-Norris and Chair Bauer-Kahan and honorable Members of the Committee. My name is Leuwam Tesfai and I serve as the Deputy Executive Director for Energy and Climate Policy at the California Public Utilities Commission.

Thank you for the opportunity to present today on the work the California Public Utilities Commission is doing to address the impact of data centers on California's energy system, especially to protect ratepayers. I do have a few slides. There we go. All right, we can go to the next slide.

So, as you know, I'll be on two panels today. And so my presentation right now will be focused on the grid impact, so we can go to the next slide. So this slide is really focused on different types of connections that we have to the grid, in particular the distribution system, but also the transmission system.

And so when we're talking about interconnection, we are typically talking about resources like generation, getting connected to the grid. And so this can be large scale, utility scale generation. But also we have, of course, smaller types of generation that are connected behind the meter for customers.

And so when we use the phrase interconnection, that's what we're focused on. Most of this presentation is going to be focused of. My presentation is going to be focused on energization.

So getting customers connected to the grid, whether that's a home, an electric vehicle charger, or in this conversation today, really, data centers and getting them connected to the grid. There are a number of different tariffs that govern getting data centers connected to the grid as well as all different types of customers.

And so as you heard earlier from Director Gutierrez on energization, when we're focusing on getting customers connected, we're Talking about Rule 16, which governs the service line connection to the specific customer. We're also Talking about Rule 15, which governs the distribution line connections that are coming to customers.

And then we also have specific rules that govern the energization of things like electric vehicle chargers. Next slide, please. So this slide is focused on taking a step back to 2023, when Governor Newsom signed Senate Bill 410 and Assembly Bill 50.

And these are two bills passed by the Legislature that were focused on the crisis of energization in the state of California. And I think we had many conversations with those of you on the panel today on these issues. And so the implementation of these bills has been very successful.

And I wanted to highlight in particular for energization, we adopted a decision in 2024 that was focused on creating timelines in support of Assembly Bill 50 so that all the customers that are connecting to the grid know the rules of engagement with the utilities and that that we have clear expectations for how quickly the utilities are getting customers connected to the grid.

And we were able to see that in addition to the, the support from Senate Bill 410, which increased the budgets for the utilities to do this work, we were able to see an improvement of over 90% from all of the major utilities and getting customers connected to the grid.

And while this is not just focused on data centers, it's really agnostic to any customers. I think it's important to talk about the progress and the success of the implementation of those bills in the conversation today in continuation of those efforts. And Dr. Gleason actually mentioned this today, the topic of flexible service connections.

The utilities have had the opportunity for flexible service connections for their customers for a number of years. And I'll explain what that is.

But I do want to highlight that the Public Utilities Commission right around Christmas put out a new decision creating a suite of strategies, standardized strategies for flexible service connections that we'll be voting on next month.

And so what this means is, as you know, to get a customer connected to the grid, sometimes you have to make upgrades to the system. That construction takes time.

But through this standardized suite of flexible service connections, customers know that they can get connected to the grid more quickly and maybe not have the full level of service that they were expecting, but they're able to get connected quickly. Their business can start collecting revenues as the utility does that construction onto the system.

This also supports the state's goals around demand flexibility as well by creating these flexible service connections. So we're very proud of this work. And again, it comes forward as part of the work related to AB50 and Senate Bill 410 to push forward these important issues of energization. Next slide, please.

So I wanted to highlight a specific proceeding before the Public Utilities commission today called PG&E's Rule 30 tariff. So I mentioned a couple of other rules earlier. 151629. So this is a brand new one from PG&E. They're leading the way.

And this was an application filed in November of 2024 before the Commission because PG&E was seeing a lot of large loads connected to the system. In particular, data centers.

And so the commission took action last to adopt an interim tariff which really created the rules of engagement there to get these new data centers connected to the grid more quickly. And I'll be talking about a few more of those details in my next presentation.

But in particular, what I wanted to focus on here is that we're getting these customers connected to the grid, even at the transmission level for these really large data centers. And so by having these standard rules of engagement, we're able to create some certainty for those businesses.

In addition to that, I want to say that this is not just for data centers, it's for large loads. So that could also include charging for freight, charging large scale electric vehicles as well. So next slide please. So I wanted to also take a step back and talk a bit about our resource planning.

So this is dovetailing on Director Gutierrez's presentation. And then of course Mr. Miller from CISO is going to talk more about this as well. But talking about the Public Utilities Commission's role and so on this slide we're really highlighting the success of bringing new resources online in the state of California.

So you can see from the slide that we have brought since 2020, over 31,000 megawatts of new resources onto the grid. Talking about the year over year success. Last year we also brought almost 7,000 megawatts of new resources to the system pursuant to the Public Utility Commission's orders and the integrated resources planning proceeding.

And so you can see that we have, we're bringing a lot of new resources onto the system. We have approximately 20,000 more projects that are in the pipeline for the near term. And so I don't just mean hypothetical projects. These are projects that have contracts or are under construction.

And so we do have a great, very strong pipeline of projects that are coming online in the very near future. And again, this is by using the demand forecast created by the California Energy Commission. Next slide please.

So I also wanted to highlight the integrated resources planning proposed decision that is actually going to be voted on next month. This is an annual decision that we vote on in order to provide the resource portfolios to the California Independent System Operator for their transmission planning.

But the commission also took additional action in response to all of these increased electricity needs to also add an additional order for new resources to that decision as well.

And so you can see from the slide that that proposed decision is going to require an additional 6,6,000 megawatts or 6 gigawatts of new resources to be procured by the load serving entities. So that's the utilities like PG&E, but also our CCAs, our CCAs and other load serving entities like ESPs as well.

And so everyone, all of the utilities and load serving entities are working together to bring on these new resources.

But in particular, I want to highlight that in that proposed decision it'll be an additional 2 gigawatts by 2030 and 4 gigawatts by 2032 because we're really trying to be responsive to this increase in electricity needs that are being shown as coming online in the near future.

And so this is not, this is not something that we take lightly. We recognize that this is very challenging work for the load serving entities, but we also recognize that there are these key needs that are are coming before us in the near future.

And I just want to highlight also in reference to Dr. Gleason's presentation, we have been leveraging globally leading stochastic modeling as part of this resource planning that was mandated by the Legislature. And so a lot of people are following in our footsteps of that work. Next slide, please.

So this is my final slot, my final slide for this part of the presentation. And I just wanted to highlight the transmission permitting piece. So as you know, the CAISO does the transmission planning process, but those projects come back to the CPUC for their permits.

Some, some of them are certificates of public convenience and necessity or permits to construct. This is also where we go through our CEQA process. And so the final note I wanted to have here is that we have implemented Senate Bill 529 to update our transmission permitting process for the first time in 30 years.

In addition to that, we are actively applying these updates to our existing permits that are in place. So we worked collaboratively with the caiso, for example, on an important transmission line serving the Bay Area and serving data center loads. That was already before the commission.

And we went ahead and worked with the CAISO in order to be able to upgrade the level of line that we were going to be permitting so that we would make sure that on day one, when that line is built, it's ready to go and to support those new data center loads.

So with that, I'm happy to turn over to the next panelist. Thank you.

Thank you. And we'll turn to Mr. Miller.

Good afternoon and thank you for the opportunity to be here today and speak to these issues. We also have some slides. I'll just see those are being pulled up. Thanks. If I could move to the next slide please. Thank you.

So I just wanted to touch on, on behalf of the CAISO, I Wanted to touch on the three general areas that the CAISO is involved in in supporting large loads. One of them is obviously more public than the others. The transmission planning is one where the industry is very aware of our activities.

The other two we've often been a bit more in the background on which is supporting the large load connection processes led by the utilities as well as the development and ensuring that the standards are in place that are needed to ensure system wide reliability, which is a little bit of a different aspect for the large load connections than load connections in the past.

If I could move to. Oh, I should have also mentioned too, just a terminology thing. The deeper we get into the transmission industry, you hear more of load interconnections, but more customer facing side of the business.

Normally refers to energizations, but it's the same thing getting the connection built so that the customer can take service off the grid, whether it's transmission or distribution level. Thanks. If I could move to the next slide, please. So on the trans. Let me touch on the transmission planning aspect first.

Data centers and other large loads play a very large role in the transmission planning. As you've heard, the large loads and data centers are taken into account in the Energy Commission's forecast as well as in the CPUC's resource planning.

And that resource planning is integrated because it also includes and takes into account transmission information that we provide to the Public Utilities Commission about the likely consequences of certain resource development. Those are key inputs into our transmission planning process and that has led already to significant transmission being approved and under construction.

I highlighted four projects in particular that were approved ranging from 2022 up to 2025. Those four projects represent 3.1 billion in capital investment.

One of the projects approved originally in 2022 were actually rescoped to be better positioned to handle higher capacity, higher load growth than what was originally anticipated, and to put us on a good footing for further additions going forward. Next slide please.

On the load interconnection side, while we lead the transmission planning and generator interconnection process, the transmission owners, the utilities themselves lead the load interconnection process. Now, the end customer rates are usually established by the local regulatory authority.

That's normally the Public Utilities Commission, but can also include the municipal utilities as well that also participate in our transmission planning process. We also support the process focusing primarily on reliability because the utilities are managing the direct connection requirements.

That includes reviewing the large load interconnection proposals to ensure that they align with our reliability expectations and our larger transmission planning process. And we have a shared responsibility about making sure that the standards are in place that need to be in place to ensure reliable system operation so that other customers aren't negatively impacted. Next slide please.

Going forward, we do have a lot of activities underway, so this is not something we're just sitting on. There are a lot of activities underway. First and foremost, continuous ongoing discussions with the state agencies both about the changes they're seeing and new actions required.

We're working both with the utilities and the North American Electric Reliability Corporation on the need for technical standards. We're not totally dependent on them. We can move on our own, but national standards can also help out.

We're also monitoring the activities at the Federal Energy Regulatory Commission where there is consideration of rulemaking for large loads and for co located load and generation.

Looking at a national perspective, we're also planning on issuing either later this week or early next week an issue paper to start our own California Centric Stakeholder consultation around certain aspects, focusing for now primarily on the technical issues as we're also looking to see what FERC what the Federal Commission is looking at doing on the national level on a policy perspective.

So those are a number of things we have underway and I'll look forward to your questions. Thank you very much.

Thank you. Mr. Medeiros joining us from.

Yes, good afternoon. I don't have any slides that'll be shown here. I think you do have some handouts that will cover my comments here and then in the next section session. So good afternoon.

My name is Mike Medeiros, the Vice President, Strategic Commercial Solutions at PG&E and my team is responsible for interconnecting large load customers across our service territory, including data centers, large EV charging centers, universities and government facilities, and advanced manufacturing customers.

We appreciate the opportunity to discuss how California can plan to grow the plan its electric grid to reliably and efficiently accommodate once in a generation load growth while advancing affordability and clean energy goals. PG&E appreciates the opportunity to provide testimony on the growth of data centers in California.

Our position emphasizes three priorities enabling reliable and timely interconnections that encourage development in our service territory, reducing rates for existing customers while protecting them from undue cost shifts and supporting California's clean energy goals. We recommend policies that balance economic development with affordability and sustainability, including streamlined interconnection processes, equitable rate design and voluntary clean energy programs.

California is experiencing broad based load growth for the first time in decades. According to the CEC, statewide electric demand is forecasted to increase by nearly 50% over the next 20 years. This growth is driven not just by data centers but by electric vehicles, building, electrification and General economic growth.

Data centers alone are projected to add approximately 4,500 megawatts of new demand, much of it offering 24 by 7, and EVs are expected to add twice that demand.

Within our service territory, we have received more than 10 gigawatts of new data center load interest over the next decade, with over 1.4 gigawatts already in final engineering and targeting energization between 2026 and 2030. These projects are often locally clustered, amplifying the need for coordinated transmission and resource planning to respond efficiently.

PG&E proactively moved away from one off interconnection reviews towards a cluster study approach. By studying large load applications together, we can identify grid optimal locations, reduce redundant upgrades and accelerate timelines for customers while protecting system reliability similar to how the CAISO plans for new generation resources.

We have completed two large cluster studies to date which, combined with additional applications, has resulted in supporting approximately 10 gigawatts of data center interest in our pipeline forecasted through 2030. Electric Rule 30, which PG&E is proud to have developed and proposed to the PUC, is a critical piece of California's grid planning framework.

It establishes clear interconnection requirements for retail transmission level large load customers and assigns cost responsibility for interconnection facilities without overloading a single proceeding with rate, design, procurement or operational issues addressed elsewhere. Importantly, Rule 30 does not attempt to solve every large load issue at once.

Instead, it provides clarity where the grid first intersects with new load the interconnection. That clarity enables utilities, regulators and customers to plan generation, transmission and permitting activities with more certainty. Grid planning for large load customers requires alignment across multiple agencies.

PG is actively engaged with the CPUC, the CEC, the CAISO and the FERC on transmission planning, cost allocation and load forecasting through process such as such as the Integrated Energy Policy Report Report and Integrated Resource Plan. FERC has already opened a proceeding to examine large load transmission issues including CO location, generation and operational requirements for reliability.

It is essential that these agencies remain coordinated with utilities and customers. Attempting to resolve all lords all large load questions in a single proceeding would delay outcomes for years. A phased coordinated approach allows California to move quickly without sacrificing reliability or transparency. PGE is also advancing flexible interconnection solutions to accelerate grid availability.

Our FlexConnect pilot allows customers with controllable loads to interconnect sooner by coordinating demand under constrained hours. Five sites have already been expedited through this pilot with roughly 80 megawatts of additional load. Load Flexibility Planned because the transmission system is built with significant redundancy, there is often unused capacity during most hours of the year.

In fact, our Systems only use 40 to 45% on average throughout the course of the year. Using that headroom responsibly while requiring customers to reduce or offset load during peak or contingency conditions can significantly shorten interconnection timelines while maintaining reliability.

So in conclusion, California's grid can support this historic load growth, but only with deliberate planning, clear interconnection rules and customer participation in grid management. Electric Rule 30, paired with coordinated planning and flexible load solutions, provides a scalable and reliable path forward that supports economic growth while protecting existing customers and advancing California's clean energy goals.

Thank you and I look forward to answering your questions.

Thank you. We'll now turn to Nico Nico Prokos. He's joining us from Silicon Valley Power. Thank you.

Good afternoon honorable Committee. Nico Prokos, Director of Electric Utility for Silicon Valley Power. It's a pleasure to be here before you today. I also have a presentation.

I don't know if it's being loaded. Should I just go or.

And I think Members do have hard copies of slides. So you can go ahead and get started while we bring up the presentation on this slide.

I'm going to start on the slide about introducing Silicon Valley power. We are approximately 130 years young. Established in 1896. We're the third largest municipal utility in California. We are the largest municipal utility in the ISO footprint. 746 megawatts of peak demand. 235 dedicated and extremely skilled employees. We are a Department, like many municipal utilities in California.

We're a Department of a city, the city of Santa Clara. Our service territory is not huge. 20 square miles. I have some slides though that kind of highlight some of the the energy density situation that we're dealing with as we've been in the data center business for quite some time. 130,000 residents, about 61,000 accounts.

And we have approximately 60 data centers currently in operation in our service territory. The next slide gives a depiction of the city and the location of our data centers. Those are the buildings in blue. The larger they are, the larger the data center. Our data centers are typically in the 50 to 100 megawatt range.

So we don't have the large data centers that you hear about on the East Coast. 500 to 1,000 megawatts. Most of ours are cloud focused data centers. They're either co locators or they're affiliated with some of our businesses like Nvidia and Intel. And I'd be remiss if I didn't point out the north of that slide.

You do have Levi's Stadium with. Some of you may be aware that we are hosting the super bowl in about a week and a half. Unfortunately, the 49ers are not in that game. But we will still go forward.

The next slide is our historic load growth slide and really starting on the far left about 2000 and moving to about 2025. So over a 25 year period you see quite an increase. A lot of municipal utilities and other utilities have been experiencing flat or even declining load growth for quite some time.

We are kind of the poster child for the opposite of that. We are seeing some significant growth in the past and we are projecting for that to continue. And I'll get into that in a little bit more detail. 55% of our power use is from data centers and the revenue increased about 14% from 2024.

I will say from a financial perspective, our 10 year projections are really showing a healthy growth in our, in our business. And again, you know, just to emphasize, we are a municipal utility here, so we're not for profit. We're here for the benefit of the businesses and residents of Santa Clara. This is one of my favorite slides.

This is the energy density slide and it shows a comparison between us, Palo Alto and the city of San Jose. Really what I want to focus on is the last part of that table, the energy intensity. And we're about 37 megawatts per square mile.

Palo Alto is about 6.5 and San Jose, which is served by the CCA PGA is 6.8 megawatts. So and this is largely driven, almost exclusively driven by data centers. Neil spoke a little bit earlier about the transmission expansion process. We are very thankful for our partnership with the ISO and also with Cec.

I think it's been an iterative process over many years. Just want to emphasize that we've been in this game for a long time and we've had lots of conversations with the CEC and with the ISO and you know, is this load going to show up? How do you model this load?

And I really feel like with that partnership we've kind of perfected that approach. And the end result, and I'll get to that in a second, is we're seeing some real tangible projects being built and being proposed in our area. Lastly, I just want to emphasize that it's really a multi year process.

You know, you start in 2020 and then you might see some projects and things proposed four or five years later. So it takes time for this. But from our perspective, the TPP process is t pppphe way to go. We've been following that process for a long time. It has a proven formula, it's resulted in projects.

I will say that we've been in this business for a long time, but in the last six months or so the amount of interest and focus on data centers has really grown exponentially. And we're seeing there's a lot of emphasis and stress on getting projects done and accomplishing it.

And from our perspective we understand that, but we also want to make sure things are done right. I think Neil touched on and others did about reliability. That's one of our biggest concerns is the reliability impacts of these projects. I don't think it's as simple as well, you know, there's spare capacity on the transmission line.

You know, we see impacts that you could have an outage on a line in Alberta and we see a voltage drop on our system. So it's all an interconnected system. And you're adding all these megawatts on there and you could have these unanticipated consequences in many different areas. The next slide is powering the South Bay.

And I'll start at the top there. This is the LS power line, which I believe the CPUC touched on, which we're very thankful and we're participating in that proceeding and we're aware that it got upgraded and we appreciate that as well. And it's interconnecting with one of our receiving stations at nrs.

To the south of that is another project from Metcalf to San Jose B. And then in the middle part, that purple hash line, that's another one that I believe is under review right now, or there's been submittals to the ISO.

And you know, I'll add that I've been in this business for quite some time and I've never seen transmission kind of be proposed and move forward as fast as it's going.

And I think that's the result of the, the partnership that we've had with the ISO and with the CEC and others in terms of, you know, what is that load growth going to look like? You know, how do we evaluate the requests coming from data centers? I mean, data centers are similar to other customers.

They come in and they, maybe they ask for, well, we need 100 megawatts. Do they really need 100 megawatts? And we have ways where we discount and evaluate that. And we've, I like to say we have lots of data on data centers.

So we're, we're, and we're happy to help and you know, in any way we can to help with that information. The blue line in the middle is our new 115 KV line. And then that, that battery, look, looking object there is. We have a new 40 megawatt battery going into service probably in the next three to six months.

And I'll close just talking a little bit about what we're, what we're doing in our service territory to accommodate all this load growth.

So it's not only working with the ISO and others to get transmission lines to our, what we call our receiving stations, it's also spending money within our service territory to be able to move those electrons around your service territory to not only to provide the service to those customers, but also in a reliable manner.

You know, when we go through and we evaluate transmission projects that are being proposed that we've requested, there are impacts to your system and then you have to make investments to be able to accommodate that so this project, and actually I should add that this is out of date, 460 million.

With this latest transmission line, that's probably gone up to about 570 million. But it's really designed to double electric load capacity to meet current and future demand. I mentioned earlier, our peak is about 740 megawatts. We're projecting to go to 1,300 megawatts in five to 10 years. That's almost unprecedented in the electric utility world.

And this growth is driven by increasing demand from large industrial customers, namely data centers. And this demand stresses both our local system and also the bulk electric system. And I mentioned earlier that most of our data centers are in the 50 to 100 megawatt range. They're cloud based, they're not AI.

Some of them, we're starting to see some interest in switching those over. And they have different profiles in terms of how they impact the electric grid. And then since 2021, we have been communicated limiting capacity to these customers. We used to provide will serve will serve letters. We stopped doing that.

We were issuing system limitation notices and we have. Now we go through a process where we do system impact studies and we have formal what we call substation agreements with these customers, which typically will have a ramp over time. And those are based on our evaluation of what our system can accommodate.

And we're in ongoing conversations with our customers because sometimes those ramps don't work for them. Or if they're a co locator and they're courting a customer on the other side who wants to lease space, they need some sort of certainty of where that's going to be in four or five years. And we work very closely with them.

And we're working on a lot of innovative projects right now, not only with the data centers, we partner with the coalition and with some of our customers in terms of how can they be more efficient on their side, how can they move that power around? Maybe between facilities latency is a big issue.

Less of an issue with AI, but for us, that's why they all located here. Our rates, I'll get into that shortly. But our rates are some of the lowest in the state. But they also need that fiber connectivity.

But being able to look into creative ways where we can potentially allow them to ramp up, but they provide flexible load. Sounds like that's something that PG&E and the CPUC are working on.

We're also partnering with other entities like Grid Care and Emerald AI to be able to find creative solutions to this that might mitigate some of those concerns regarding some of the costs related to data centers. And with that, that's all I have and I look forward to your questions.

Thank you. All right, we'll now hear from our. Last speaker for this panel. Karib Bunder is joining us from the Data Center Coalition. Thank you.

Thank you. And I believe I also have some slides for this portion of the presentation today. Thank you, Committee chairs and Members of the Committee for the Opportunity to speak today. And thank you for the thoughtful time that you're dedicating to this critical issue. As mentioned, my name is Kara Boonder.

I'm a Director of State Policy at the Data Center Coalition. I myself am based in Washington, D.C. and greatly appreciate being out here in the sun and warmth compared to what it is like out there.

So as the slides are getting pulled up at a high level, just to give you an idea of who DCC is, we are the national level trade association serving as the voice of the data center industry, primarily focused on advocacy and education on behalf of the industry.

If you can go to the second slide, please, that would be great. So we advocate for business climate policies and investments that support the growth of the data center industry and the competitiveness of the industry. And we're also here to serve as an information resource for elected officials, stakeholders, candidates, community leaders and others.

I think we are seeing that we've gone from a time where it was what's a data center? To what is a data center? And why am I hearing about it so much? So that's why we're here. If you can go to the next slide.

So we represent leading owners and operators of data centers, in addition to those who lease large amounts of data center capacity. And we've defined that as over 50 megawatts. You go to the next slide.

Part of my presentation today is going to I know this is supposed to be focusing on energy and AI impacts, but I think it's also really important to consider that data centers are not only growing in number because of AI, there is still a lot of demand being driven by our own consumer use.

And I think this is one of my favorite stats to present, and this comes out of Deloitte, that the average U.S. household has 21 connected devices.

So if you're thinking about everything from your computers to your printers to your smartwatches, I just bought a smart litter box because I travel a lot and I think my cat appreciates that.

But I think that just gives you an idea how sweeping the products and different purposes that are served by data centers that are processing that data and facilitating the even mundane things that are part of our daily lives. The other component of this is we're seeing more and more of the global population coming online.

And I think the cloud almost becomes a misnomer because it turns out that, that it's not actually in the cloud. It has to be stored somewhere physically. And that is a data center. Can you go to the next slide please?

And so this is another way of looking at this increase in consumption just based on average household monthly broadband, broadband consumption. And we just see the exponential growth that's occurring there. Next slide. And then this is another way of looking at it between commercial versus residential demand drivers here.

So I think it's kind of surprising to see that the residential demand is outpacing the commercial demand at this point. Next slide please. And this is again going back to contextualizing the demand growth that's coming from AI versus some of the more traditional cloud based services that are being supported by data centers.

AI will become a growing portion of the demand drivers for data centers, but we're not quite there yet. So and I think we'll also see, you know, how that growth comes online in different ways also will shape the types of things that can shift on and offline.

And when we're talking about things like demand response, if we're having cloud services, it's not as easy to turn those off because then, hey, we don't have access to those things. And then that makes, you know, the financial transactions, the E commerce that we depend on, our telehealth appointments, our emergency services, we need access to those 24/7.

And that's why data centers do require that significant uptime and reliability. Next slide please. Again, this is another way of looking at it. As you can see, that light blue portion of the graph is where there's all the demand for data centers, excluding AI.

And then that top portion is where you see that bifurcation for the traditional services provided by data centers versus the growing proportion that is attributable to AI. Next slide please. I just also wanted to provide some context in terms of the primary data center markets across the country.

So this graph here shows current data center inventory versus that that's planned or under construction. So of course you're going to see Northern Virginia, the largest data center market in the world over here on the left.

And then you can see some breakdowns of other markets as you move further down to the right side of the page here. We're seeing a lot of growth in markets such as the Atlanta metro region in Texas.

And you can see, I can't see from where I'm sitting which bar there is Silicon Valley, but it is one of the ones further to the right here and you can see an even smaller portion that's attributable there for planned construction.

That's not to say that this is not an important market, but I think you are also just seeing a greater shift to other areas of the country.

And of course there are a lot of citing considerations that go into that land costs, electricity costs, workforce, the regulatory and tax environment, more broadly, a variety of different things go into that. But just wanted to provide some context for where we're seeing that projected growth for data centers across the country in terms of primary markets.

I also have data on secondary and tertiary markets as well and would be happy to follow up with that. The next slide please.

So as the data center industry is committed to ensuring that there's a resilient, affordable and reliable grid for all customers, we certainly need to rely on the grid very heavily because again need to have that uptime to provide all those critical services that are supported by data centers.

I do think it's also important to consider that while data centers are large consumers of energy, there have also been a lot of efficiencies gained from moving from the more traditional older bespoke having on site server closets that needed to be managed and cooled in those individual facilities to having them concentrated in these larger campus style buildings that host multiple servers.

Whether that be for an individual company or for, as mentioned earlier, the co location facilities that Lisa Server space to other customers. So as you can see from this slide, by 2018, 80% of data center computing took place in these larger utility scale data centers.

And while there was a while the computing output jumped 550%, we've actually only seen energy consumption rise 6%. So I think that really speaks to the efficiencies gained by having these housed in these larger facilities and taking a more, I guess, deliberate approach to how that is managed. Next slide please.

Also want to talk about impacts on AI and the grid. I think this largely is framed typically as a taking from the grid for data centers, but I think we also need to keep in mind that there are efficiencies gained and some innovative solutions that are supported by data centers coming online as well.

Some of those include smart thermostats allowing us to have a better idea of our own energy consumption in our homes and tailor that to different times of day or whether we're there or not.

Also for managed EV charging Understand a lot of the electrification coming online here in California, but also grid enhancing technologies that allow us to really leverage the existing infrastructure that we have rather than having to invest the additional money for contributing to new infrastructure.

So I think it's important to keep that in context as we're talking about this as well. I don't have a slide for the rest of what I want to cover today, but wanted to be responsive to some of the topics that have come up earlier.

Certainly understand that load forecasting is a challenge, especially as we're seeing this unprecedented load growth.

Of course, not only attributable to data centers, but you know, onshoring and reshoring of manufacturing, hydrogen production, electrification, as mentioned earlier, and who knows what the next new innovative technology that we haven't even thought of, what types of energy demands will be associated with that.

But the data center coalition, for our part, our energy team has been really leaning in as an engaged stakeholder with grid operators, utilities, other stakeholders, understanding that this is a really, really critical moment to meet and do understand that there are some complications associated with our data center low growth.

So just wanted to run through a couple of the approaches that our energy team has put forth in terms of trying to build greater confidence in that forecasting by identifying certain tools and best practices that can be employed. So we often hear about, you know, is this growth real, is this speculative?

So one way that we would suggest that getting at that is examining the commercial readiness of a project, whether that means looking at executed service agreements, contracts, other things that might indicate that the project is moving forward. Where are they in the permitting process, is there, have they already put down collateral?

Also trying to establish more transparency in the forecast assumptions to understand how those forecasts are being decided upon and put forth, trying to establish some type of reporting standardization across transmission operators and electricity distribution companies, supplementing forecasts with third party data, doing some scenario planning, but also doing regular forecast review and backcasting to take a fine look at how have the past forecasts turned out in terms of reality and being able to use that as a benchmark to guide future forecasting efforts.

So I look forward to continuing the conversation today and offering some additional insight on behalf of DCC with regard to AI and rate payer impacts as well. And thank you for the opportunity to speak today.

Well, thank you and thank you to. All of our panelists for being here offering your insights and perspectives. We'll go ahead and open it up for Member questions. Any questions from the panel? Mr. Rogers.

Thank you so much.

Actually, I wanted to dig in a little bit on some of the conversation around flexible load, which 50,000 foot, as best as I can explain it to my constituents, is that the more that we can fit in terms of load on the existing infrastructure, the more it'll distribute the costs and therefore hopefully bring down their rates.

But if we have to build additional infrastructure, there is that additional cost that they're paying attention to. And that really comes into effect especially around high usage times in the summer when it's really hot.

And I heard a couple of different comments about trying to coordinate some of those activities relative to data centers to be able to piece this all together.

So my question is kind of collectively for the CEC the CPUC, Caiso, do you feel like you have the statutory authority and also the data that you need to be able to make those types of decisions? Once we do get to the point where we are seeing our rates impacted by the additional growth.

I'm happy to start. Thank you. Leuwam Tesfai Public Utilities Commission so starting first on the topic of flexible service connections, we do feel we have the authority to be able to do that. We did issue a proposed decision a few weeks ago on that issue.

And so really a number of customers have worked with the utilities on an individual level in order to access flexible service connections. So to your example, you gave a great example.

So maybe the utility is able to serve them at 100%, 85%, 90% of the year, but during these times of strained grid connect conditions, they are agreeing in advance that they would be reducing their use. And so that's been happening on a one off basis.

But by putting it in a decision at the commission, it standardizes that suite of offerings for any customer that wants to be able to access that. And so that helps them with their business, the certainty.

But then in addition to that, and I have this in my second panel legislation from last year, SB 57 is going to be having us do additional research and study about these ratepayer impacts.

And I think that that is going to help us assess if there are any additional authorities needed to be able to take these opportunities to the next level. I turn it to the other panelists if they wanted to add anything.

Yes, it's Neil here.

I actually would like to jump in just to help also help clarify something because the flexible load discussion, often in a number of conversations we hear it used to be referring to two different things and it creates a lot of confusion because some of the examples that Mr. Medeiros was referring to are about people looking at providing some Additional flexibility up front either so that they can get connected more quickly by taking, I'll say a lower level of wire service and also that some are also looking at the possibility of on a more permanent basis, taking that lower level of wire service to reduce their cost of connection.

But that's all about the wire's capacity into the area and so on.

Now the other part about the flexibility is if there's flexibility available to participate in the day to day service, overall supply, demand, balance on the system about is there enough energy, what's the way to lower the total cost of delivered energy to customers day by day, hour by hour?

And if there's flexibility that those data centers can provide to back off during high priced periods, that's also another important attribute. It's not in our system really necessary just to get connected sooner, but it's an important attribute as well. But in a number of jurisdictions, I.

Think I actually was meaning more about curtailment day to day as those loads are available.

And we often conflate the two. And that causes some confusion.

Now that issue, once they're connected, that's another important issue that the ISO and the state agencies are looking at about what can we do to help capture that benefit, to provide the economic framework that encourages data centers to provide the flexibility they have to help more aggressively participate in the market, keep the costs down for consumers across the board.

That is a conversation as well. It's a little disconnected from the wires in some jurisdictions. The two always get put together because some jurisdictions are having more trouble getting new capacity built than actually we are. We do have the benefit of the integrated resource planning process to drive new resource development, but a lot of other jurisdictions don't.

So they're using the interconnection as a lever to put pressure on providing that market flexibility as well. But both are important conversations.

We do see the opportunity there and that's something that we have to work with the state agencies on to make sure that we actually capture and get to take advantage of in the energy market, both for cost and for reliability during those hot summer days.

If there's flexibility there from a system wide perspective, that can help pull the demand down, that helps save capital cost and it also provides higher reliability. Those are things we're really concerned about and want to explore.

And I think the second part of that question is key, and I haven't heard an answer on it, which is if you have the statutory authority, and it sounds like you think that you do, to be able to have those conversations and help to start to design that system.

In the event that we get to that point, do you have the transparency and the data to be able to make those informed decisions?

Yeah, I'll speak on that. So from the demand forecasting perspective, we feel that we have sufficient data through what is being provided by the utilities. And I will take this opportunity to thank the utilities for their collaboration in preparing this year's demand forecast and helping us think through some of these challenging issues.

But we received the data interconnection application, data center interconnection applications and I mentioned some of the information that's coming through those. And then we also are getting ramp rate for each data center from some of the utilities.

And then additionally we have customer interval meter data from utilities which we can use to develop typical data center hourly profiles for the peak demand forecast. So regard, I think we have sufficient authority. We're getting the information that we need.

There are some new challenges and uncertainties that come up with, you know, getting a laundry list of new data for these projects. So we, we feel that we currently have what we need to prepare the demand forecast and do the integrated and proactive planning.

Yeah, if I could. Yeah. I think there, there is more work to be done around load modification, load management for data centers. There's. We don't want to be prescriptive because I think there's different opportunities for them to participate and actually be an asset to the grid.

Whether it's reducing their compute load at one location and moving it to another state, if they have enough time visibility to do that, whether it's just traditional load management on site, lowering the temperature, or whether it's maybe, maybe they don't have the ability to do that.

But can they pay other customers to reduce their loads during those time frames? There's a lot of different things that we can do.

I think the key is to make sure that it's operationally reliable, that our operators, our grid operators can rely, that it's going to show up in the time frame and at the, at the level that they need to make sure they maintain reliable grid. So that's going to be a key thing.

I think we're all going to have to sort of double down on.

And if I could just add, just to emphasize that, that I mean we have that capability right now. It's at a smaller scale. I do think the, the flexible approach. I think there's broad agreement that that is a good approach and we're all evaluating that. I don't think that it's a replacement for the TPP process.

I think that that's a successful model. So that really some thought needs to go in there.

I think some of the challenge that you have talking to the data centers is that, for example, if they're going to be running and what we're talking about is they're not curtailing, they're going to their backup generation and that's going to run more. And you have certain areas like our area where there's air quality limitations.

And so they have some concerns about, well, when we talk to them and we're working through that process right now is how much are we talking about? And then are we going to be running into a challenge necessarily with the air quality district? I think this is, it's a tool in the toolbox.

But the numbers that I'm seeing are big numbers that they're throwing around. And if all that starts to show up and again, I'll use that Alberta example, it's like you could have something that happens. And we see it now, we saw it last summer and it was a mild summer.

You get into a hot summer and you're adding 1,000 megawatts onto the system, there's potential for problems to happen, even if you have that ability to. And like I said, we have that now to be able to do that.

The only thing that I would add is what we can do right now is more of a, what I call a blunt instrument. So if we run into a problem, we can drop a feeder. You might have three or four data centers on that feeder.

What we're in conversations right now is more of a surgical approach where you might just drop one out of the three, not all three. And that's sufficient to address the issue.

And then also that addresses the problem of some of the air quality concerns because instead of having three data centers now on their backup generators, now you only have one. And that's the benefit of that program. So there's a lot of pros and cons there, but I just, I think that it's one of the.

There's a lot happening right now and it just needs to be methodical and thought out.

Please go ahead.

I was just going to actually circle back to say I realized we all jumped into the issues we're working on but didn't actually address the straight up question, which was, as you've heard, though, jurisdiction isn't the problem.

It's just a case of doing the hard work of getting the right structure in place, the right models in place, working with the data centers and understanding the flexibility they do have, so that we're developing our systems to take advantage of that flexibility and create the right incentives for them to provide that flexibility to the extent it's available.

But jurisdictional issues aren't the problem there. It's a case of, and we have a stakeholder process underway on some of this, of doing the hard work of getting the right framework in place to capture that flexibility. But jurisdiction hasn't been the issue.

Yeah, no, I really appreciate that. And I really appreciate the flag on the backup generator issue, something that I hadn't even been thinking about. But in the analysis for the informational hearing, which shout out to Committee staff was really good.

One of the things that I did take note of was, especially within the context of the AI discussion, that sometimes, like when you're training an AI model, it is exceptionally energy intensive, but then other times it's not.

And if there's a level of communication, and maybe the free market fixes this because you wouldn't want to train your AI model when utility rates are higher. But if there's a level of coordination that even looks beyond just what each individual data center needs.

But when companies are doing these energy intensive projects that they know are going to be a huge strain on their available resources, that the communication and the coordination is happening. So thank you so much for the. Oh, yeah, please.

Can I respond to that just to add a little bit more. So the Public Utilities Commission this year is opening an advanced rate design rulemaking. And so it talks about some of those things you're talking about where maybe we can use price signals from the grid in order to incentivize customers to take different actions.

So when you're talking about training models for AI, if there are certain price signals that are in the market that companies can take into consideration to decide when would be the best times to be able to do that. So we have existing authority for that and are opening that rulemaking this year.

Excellent. Thank you so much.

Assemblymember Schiavo.

Thank you so much. So much going on. I was very interested to hear from you, Mr. Procos. Is that right? Okay, Mr. Procos, because you have a lot of experience on this where, you know, it's a new thing in a lot of areas.

And I'm curious if you can expand upon one of the things that you said you're looking at. Well, I guess this is the next panel. I don't know if I'm going to. Be here for the next panel, though. Okay. Oh, I'll hold it. Okay. So hard. I know that's good. Silver Spring, Right.

So one of the things that you talked about, I'm curious if, because what we've been talking about is kind of smaller scale, a lot of smaller scale data centers, it sounds like, are mostly in California.

And I can imagine all kinds of reasons for that land being one of the big challenges and environmental regulations maybe being another one and other regulations. But have you seen that? Because we just had this discussion on reliability. Have you seen that?

Do they ever upscale or do they always kind of come in and use the same amount that you expected, or they're pretty consistent, or is there space for them to grow once they're established, and then it creates more demand that you don't anticipate?

And also, I was hoping you can explain a little bit more about you were saying that clouds versus AI have different demands. And can you explain that a little bit more.

Certainly. So, I mean our experience with the data centers that we have in our service territory is, I mean again, like they're not unlike a lot of large customers. They come in and they might say, and I'll just use an example, they need 100 megawatts. We tend to discount that.

I mean for the ones that are in our service territory, I mean I'll throw out a ballpark number, but we typically see them come in around 65 or 70% of nameplate. So, nameplate would be like that 100-megawatt number and that's fairly typical for those.

So, when we go through the process where we're doing our load forecasting and talking to the CEC and we've kind of, I think reached common ground in that area, and that's been really helpful for us. I will say that in the last six months, again, there's I think so much focus on this sector.

I mean I spend a significant amount of time of my day talking to data centers. And I'll tell you, the common theme that they're talking about is they need more energy. In our service territory, we have, you know, I mentioned earlier about the substation agreements and we have different vintages.

Some of the older ones, it might just be one number. The newer ones might be a ramp. So, and we structured those where you may be using, again, the 100-megawatt number, 25, 50, 75, over a five-year period.

And that may or may not work for those customers, but for us, it's helped us kind of manage the potential impacts on our, on our grid and also, line up with the work that's happening external and bringing the transmission. But I will, I do want to emphasize that it's very difficult for them.

I mean we do hear a lot of they need more power. I mean, they're in discussions with the Amazons and the Facebooks and typically, those customers, they want a commitment that you're going to provide nameplate before they sign a lease with those, the colos, co-locators, the standalone ones.

And so, and that's a little bit of a newer thing, again, in the last 6, 12 months, where it's really kind of taken off. And you had a question about cloud versus AI. I mean, we have a lot of experience on the cloud. These are great customers from an electric utility perspective. Typically, I mean they're baseload customers.

You know, 80% utilization factor. I mean it doesn't get any better than that from a revenue perspective and kind of a cost of service perspective. The AI ones, and we have less experience in that area, but what we're hearing is that they can, they'll, they move a lot more, they'll ramp up and down.

And when you get into that kind of situation, and I'll go back to what I mentioned earlier, where, you know, what keeps me up at night is, is reliability. Right? And as you add more of these, and then maybe in California, we might get some 500-megawatt AI ones and it's like you're going to see some issues.

And I'm sure the ISO is also concerned about that. Frankly, I don't know how they're doing this in Virginia. I can't imagine all that they're adding there. And you're starting to get a lot more attention from FERC and NERC and because there's a recognition of some of those potential reliability issues.

I don't know if I answered all your questions there.

Yeah. Yes. And if I can add into the mix when you answer is, is there risk for these becoming standard assets? And I know in the conversation earlier, there was discussion about, you know, during the high demand times, asking them to go out of state or utilize other spaces. How easy is that? How quickly can that happen?

How do we handle that situation?

Yeah, I'll say on that. The ability to move compute is still something data centers are really looking hard at. So, it's something that's not infant stages. So, they're looking at it as a potential option. I wanted to get to your question. We do have some data centers within our service territory.

We see many, many times that data centers requested a certain level of capacity and they're nowhere near it. So, they have room to grow within the footprint of their data center.

In their cloud or AI?

We see cloud and AI and inference, excuse me. What I will say is, as we see that, and it's part of our Rule 30, that's why we believe that the minimum demand charge—we're going to talk about it more in the next session—is critical because, you know, you're asking us to invest a certain level of capital in the infrastructure is not being used.

So, making sure there's like a minimum revenue that we're being able to achieve every year is really important.

And if, if I may on the AI point, because I think it's an important one to make, is that the, you know, if you look at page seven of the backgrounder inference, which is basically our usage of AI is where the vast majority of energy is going. That's where they want less latency as you heard.

And so, those need to be where the users are and those probably can't respond to market cues because we're not going to respond to market cues. Right? We don't know that we're doing an AI search and that's resulting in more energy costs necessarily.

And so, I just think we need to understand where the energy is coming from on the AI front and how a market mechanism would function in reality, because I don't see how that would change our behavior, which is most of what the data center energy on the AI side is coming from.

All right. Do you have another question?

I think Mr. Procos had another comment.

I was going to address the stranded asset question, but I don't know if that—should I hold on that to the?

Maybe, if we don't. Yeah, no. We're keeping everyone waiting for the conversation on rates and affordability. But I have a couple of hopefully quick questions related to our demand growth and forecasting. So, this is from the slide deck. So, and I'll first just acknowledge this is incredibly difficult, incredibly challenging.

You framed some of those challenges in your presentation, as did our panelists from Lawrence Livermore. But what we see here is that basically, like your forecast flatlines after 2035, and I don't think—no one thinks that's actually true. And I think in many contexts, a 10 year out forecast that we know is inaccurate kind of doesn't matter.

But in this space, where it's taking us 10 years to build a new transmission line, we actually do need to have more robust out year forecasts. So, I guess my question would be what is, you know, when you look at this forecast, what are you thinking?

How are we going to make the out year forecast more accurate so that we can actually plan for infrastructure that requires incredibly long lead times?

So, we're constantly updating the forecast through our integrated energy policy report process, and even I presented data showing through December 2025—it's real time data coming in and we're constantly building that into our process and methodology.

But we are, as we get new information about application status, we are modifying the confidence rate and the load we expect to see from those projects. So, I think we acknowledge that the out years are very uncertain and are constantly looking at new data sources and updates that will help us to revise and update the demand forecast.

I just wanted to follow up on this chart because I understood this—and please correct me if I'm wrong because I'm not an expert chart reader. This is different forecasts and so, your margin of error, as I read it, is 2,500 megawatts or 25%. So, am I reading that right?

Are you talking about the difference between the planet forecast? Yeah, so we use them for different purposes. I wouldn't say it's necessarily a margin of error, but different assumptions going into the different forecast builds, so, and the local reliability is using more kind of aggressive assumptions and used for the purposes that I mentioned in the presentation.

I mean, I think it's interesting because I was looking at, prior to this hearing, the PJM new rules that came out in January 2026, which, for those who don't know, is the Midwestern Energy Consortium with, you know, states such as Kentucky and Virginia.

And so, we're talking red states, blue states, and they came together and set these rules.

And one of the rules that, as I understand it, was you had to put money down when you made that tag number so that we had more certainty in what people were actually asking for so that we were better off in the long term as we understood where we were going.

So, I guess, like I'm hoping we're looking at something like this because I don't see how this kind of differential can lead us to enough certainty to do good planning.

If I may respond. And I'm sure Neil has something to say about this as well. So, because, as Director Gutierrez was saying, you know, this work is done in a cycle, and so, we are able to be very nimble.

And so, for example, the slide that I showed, I think it was number seven for me, in that presentation, I was showing in the IRP decision that's on the agenda next month, we're adding an additional 6,000 megawatts for the near term. So, that's 2 gigawatts by 2030, 4 gigawatts by 2032.

And this resource planning process that we have, that's very tight linkages between CEC, CPUC, and...allows us to be nimble. So, we are extremely different from PJM, which has struggled through their capacity market to bring new resources online. They have not had the right signals in place to be able to get that new generation built.

And as you saw from the slide number six that I had, we have not had those challenges.

And that's because of the cycle planning process that we have every year. We are able to be very responsive to these new load numbers and be very responsive to create signals to be able to get new resources contracted, not just because we have a signal, but because we order the load serving entities to do it.

So, Neil, is there anything you want to add there? Because we're quite different from PJM. We are—yeah.

The other thing I would just add is that when we are doing the transmission planning, especially into a tighter area like, and I'll pick the Greater Bay Area, we do want to use a higher level of a more aggressive forecast that's all within the same family of assumptions, but go to the higher end of the range for the forecast within the local area because now you're not only forecasting the total volume of load, but where.

So, when we're looking at planning into that area, we always want to be a bit on the high side, be more aggressive, because we can always slow a project down, but it's very difficult to speed it up.

So, there's a deliberate strategy on the system wide forecast and of using a bit more moderate set of assumptions looking system wide. But when you're looking into a local area, we want to aim at the higher end of the range of the possibility of the forecast so that we're not caught short.

And if we are early, like I said, it's a lot easier to slow a transmission project down if you think you're coming in a little early, that doesn't usually happen. But if it does, you can slow it down, but it's very hard to speed it up. So, that is why we're marked by conservative.

We're looking at the higher end of the range so that we're getting the transmission going because it is challenging to get these major projects built.

And in her comments, Ms. Bunder from the Data Center Coalition outlined some best practices for forecasting methodology including, including commercial readiness, back casting, and evaluating performance against previous forecast to inform assumptions moving forward. Are these things that we are doing as part of your forecasting methodology?

So, we are planning to track application status as I mentioned in the previous response, and that includes which data centers are completed. So, we will pull all of their advanced metering infrastructure data for those projects and apply them in the—in our forecast revisions.

Any, anything else? No? Okay. All right, so my last question before we move into our next panel is related to the impact of FERC's ANOPR. What do you—what is the impact of that on California?

I can start on the ANOPR. So, the public—California Public Utilities Commission—serves as the representative for California ratepayers before the Federal Energy Regulatory Commission historically. So, that was me in a past life. I used to be one of our attorneys appearing before FERC. And so, that ANOPR that was instigated by the Department of Energy, we did file comments there.

And so, there were a couple of different parts that we filed comments to. And I think Kaisal will want to jump in here in a moment as well.

But first, I want to start with is there is a number of disagreements between the states and the Federal Government about the level of jurisdiction that FERC is infringing upon. And that was clearly demonstrated in our comments.

So, you know, through these presentations today, we have shown you how California in particular, but other states as well, have decades of experience with managing energization. And we're also moving forward with, for example, PG&E's Rule 30 application on new, advanced ways of being able to bring these new resources online.

And so, we made that very clear in our comments. That being said, you know, if there are opportunities to find improvement at that bulk service level, you know, we provided an opening there to work with the federal regulators on that topic. And so, we're actively participating in the ANOPR. Neil, did you want to add?

Yes, I would say largely the same thing that we are focused on this was when we were participating in that process, we were providing, I'll say, more of the fact base of where are the issues that we see and what actions are moving forward to address those concerns. The ISO is not getting involved in the jurisdictional issues.

We saw that to the extent the state of California took a different position from what was laid out in the ANOPR that the Public Utilities Commission would be looking at those issues. Now, we do see that there are parts of the ANOPR that do directly relate to system reliability concerns.

So, there was a whole range of issues identified in the notice of proposed rulemaking. Some of the issues that are, I'll say, clearly within the ambit of reliability on the overall transmission system, we saw that some of those seemed reasonable areas to pursue.

But at the end of the day, we're having to wait to see what FERC does coming out of the ANOPR and then we'll respond from there.

All right. All right. I think what we'll do is go ahead and wrap up this panel so that we can move into our next panel, which is focusing on the impact on ratepayers. And as I said, there are many of our current panelists who will be staying for the second conversation.

So, Executive Director Tesfai will be staying with us, as will Mr. Medeiros, Mr. Procos, Ms. Bonder, and then we're going to be joined by Karen Hietta from the California Public Advocate's Office. Thank you to our panelists who are departing. We appreciate you being here. All right. And I think we are beginning Ms. Tesfai with your testimony.

Thank you. And I have two slides, so I'll wait for those to be pulled up. Also, I realize we have the clicker here, so we could actually have been in control of our slides the whole time. So, I will use that this time. All right, so focusing, next slide.

Oh, I don't have to say next slide because I can click this. Okay. So this slide deck is going to be focused on one of the most important aspects of the California Public Utilities Commission's jurisdiction, which is the protection of ratepayers.

So, my presentation will be focusing on the opportunities that we have to be able to look at how costs are being shared and allocated between different customer types, as well as highlight some implementation of recent legislation that we'll be doing this year.

So, turning back to the PG&E Rule 30 application that is before the Public Utilities Commission, I mentioned earlier that we adopted an interim decision in 2025 that was really focused on approving the structural components of Rule 30 for customers to be able to get connected to the grid more quickly.

But we did not adopt a decision on the cas—the cost allocation. And so, that's what I'll be talking about, what we're starting to look at here.

And so, on this slide, what we're really highlighting is that through this application, PG&E is looking at different reimbursement processes for different types of transmission, different types of upgrades that are done for a customer to get them to be able to be connected to the system. And so, there are different types of upgrades there.

So, we usually call them type 1, 2, 3, and 4. And so, types 1 is the transmission service facilities, type 2 being the transmission interconnection upgrades, and type 3 being the transmission interconnection network upgrades, and type 4 transmission network upgrades.

So, focusing on transmission service facilities, interconnection upgrades, and interconnection network upgrades, the application, what we're looking at is creating this opportunity for the customer to be able to advance those type of upgrades for the large load to connect, this being a data center or some type of heavy duty vehicle charging, for example, and being able to do that work by advancing those funds.

And in order to get the work done more quickly, some of that work can also be done in kind.

And so, these types of upgrades, something that casts them apart from the Type 4 upgrade that I'll talk about, is that it's very clear that all of these upgrades are to support that individual customer that is joining the system, so the cost allocation is a bit more clear.

And so, what we're looking at there is through the proceeding, we're going to be making sure to fine tune so that we are really making sure that that customer that is driving all of these capital costs to the system is paying their fair share.

And so, what we'll be looking at very closely, there are things like demand charges. This is not something new for this application.

Demand charges have existed for a long time and demand charges, at a high level, are the charge that is charged to a specific customer, not really residential customers, but like industrial and commercial customers, in order to build the system at the peak need for that customer.

And so, when we're doing that, you might look at a customer who's saying, oh, I'm going to do some behind the meter, behind the meter resource, I want to use some other type of generator, but they still want to have the electricity grid there as a backup for them so they're able to switch on it and operate at the level that they need.

And so, we need to make sure if the utility is doing those upgrades to support that level of need that that customer is paying for it.

Now, I wanted to highlight also the trans, the type 4 facilities that I just mentioned earlier, which we call transmission network upgrades, those are a little bit further up in the system. So, it's not necessarily the entire upgrade is directly related to a specific customer.

And so, we're going to be doing a higher level of analysis for those upgrades to make sure that we are accurately allocating those costs to these specific customers. And then, in addition to the allocation of the cost to the customers, we're also looking at the payback period for some of those advanced payments.

Eventually, those customers get those funds paid back once they are paying their revenues and their electricity bills. And so, that's how that gets paid back. My next slide is about Senate Bill 57. So, many of you probably remember Senate Bill 57.

And so, this was legislation passed last year that is requiring the Public Utilities Commission to do a specific ratepayer analysis to understand, you know, across all of the service territoryies, not just PG&E but SDG&E and in Southern California Edison as well, what type of strategies can we be employing in order to mitigate impacts to ratepayers?

We recognize, and I'm sure PG&E will talk about this in more detail, that there, and Chair Petrie Norris mentioned this as well, that there are opportunities here with these types of customers to be able to spread out fixed costs, like wildfire mitigation, for example, and other fixed costs across a larger per kil—per kilowatt hour basis.

This is also a conversation we have around electric vehicle charging and other electrification of the system. But we have to be able to study the unique impacts that data centers have on the grid in this way.

So, this Bill, and we have a BCP related to this bill that we're advocating for this session, is going to be a key part of our strategy so that we're able to come back to the Legislature and the Administration to talk about any kind of additional authorities we might need, as Assemblymember Rogers mentioned, or other maybe jurisdiction that we need in order to make sure that we're able to create these strategies for ratepayers.

And one thing I want to take a step back and say is that the Public Utilities Commission, this is our historical area of jurisdiction, protecting ratepayers doing rate design, doing cost allocation.

This is not new for us, but we are seeing that there are some unique impacts for data centers and this legislation is allowing us to study those.

Thank you. Now, we are turning to Mr. Medeiros.

Great. Good afternoon, again. PG&E appreciates the opportunity to discuss how large load interconnections can reduce rates for existing customers while ensuring customers pay their fair share of infrastructure cost. California faces unique high electric rates driven by wildfire mitigation investments, public purpose programs, and existing policy commitments such as net energy metering.

At the same time, the state is confronting a budget deficit and need for economic growth, and in this context, attracting high quality committed large load customers is not a risk.

It is an opportunity to spread those fixed costs, those fixed system costs, over a broader customer base and reduce rates for everyone while delivering prosperity and jobs and increased tax base.

Unlike many other states, California socially allocates cost for programs such as CARE, FERA, wildfire prevention, and other public purposes, and has adopted an energy metering program that significantly reduces the customers paying the electric grid cost, even when they rely on the grid.

Increasing usage by customers who can contribute significant stable revenues is one of the most effective ways to improve reliability affordability system wide. Electric Rule 30—we propose Electric Rule 30. It was specifically designed to prevent cost shifting from new large customers wanting to connect to our grid and protect the existing customers from cost shifts.

Large load customers are required to fund the cost of their interconnection facilities up front based on actual, not estimated, cost. This ensures that existing customers are not exposed to cost overruns.

Customers may either pay PG&E to build the facilities or construct them independently and transfer ownership back to PG&E once they're complete. In either case, because the customer is funding the infrastructure improvements up front, the interconnecting customer bears the risk if the project does not materialize.

Refunds occur only after the customer begins taking service and generating revenues, and only in proportion to the revenues. To further protect ratepayers in Rule 30, we've also included a minimum demand charge, minimum contract term, and early termination fees—tools already adopted in other states to avoid stranded infrastructure cost.

Some have questioned why large custom margin loads customers are not required to fund all transmission network upgrades. Those are the type 4 upgrades that my peer was just talking about. Transmission network upgrades function more like a highway than the driveway. The customers that pay, they're paying for the driveway. The network upgrades are more like the highway.

They serve thousands of customers simultaneously and broader policy goals, such as electrification and reliability. Large load customers pay transmission rates once they're energized, just like other customers, and therefore, contribute to these costs through their ongoing rates.

This approach is consistent with long standing CPUC FERC precedent and ensures that infrastructure that benefits many customers is not fairly assigned to just one. PG&E's analysis and Rule 30 proceedings shows that revenues from large load customers are expected to reduce existing customer bills.

Over a 10-year period, typical residential customers could see bill reductions totaling approximately $3,000 due to fixed costs being spread out over a large revenue base. Just to be clear, that $3,000 over the 10-year period that I referenced. While individual cost categories may shift, from a customer perspective, what matters most is the bottom line bill.

Increased revenues from large load customers help offset transmission, public purposes, and wildfire-related costs that would otherwise be borne by a smaller customer base. PG&E does not support creating a discriminatory rate classes for specific industries.

Customers that impose similar generation, transmission, and distribution cost and exhibit similar uses patterns should be treated comparably to comply with state and federal law.

However, California should use upcoming proceedings, large load proceedings, such as the anticipated large load rulemaking, to thoughtfully consider cost allocation, rate design, load management programs, and procurement issues in phases, rather than delaying the urgency needed decisions related to interconnections.

And in closing, well designed large load policy can support economic growth, protect existing ratepayers, and help stabilize electric rates. Electric Rule 30 provides strong safeguards against cost shifting, while enabling customers to bring significant new revenues onto the system.

With clarity, expediency, and coordinated regulatory action, California can turn large load growth into a meaning affordability solution for all customers. Thank you.

Thank you. All right, we'll now welcome Karin Hieta, who is the Program Manager of the California Public Advocates Office. Welcome.

Good afternoon, Chair and Members. I have some slides to go with my discussion points here, if you can bring those up. I appreciate it. I'm Karin Hieta, the Program Manager for the Energy Infrastructure Branch at the Public Advocates Office at the California Public Utilities Commission, or CPUC. I'll start briefly with who we are and what we do.

If I can get the slide to go. Thank you. The Public Advocates Office was created by the Legislature with a mission to obtain the lowest possible rates for customers consistent with safe, reliable service and meeting the state's energy and climate goals. We advocate on behalf of ratepayers, with an emphasis on residential and small business customers.

Affordability is our North Star. We primarily carry this out by scrutinizing every utility proposal that could increase cost to ratepayers, and in many instances, we provide alternative, lower cost proposals. We're honored to come here today to discuss data center issues in California.

As we've already heard, data centers are poised to power the future of AI and cloud computing, and the facilities are extraordinarily energy intensive. Individual sites often exceed 50 to 100 megawatts, equivalent to the residential load of midsize cities, such as Santa Rosa. As demand grows, utilities are planning billions of dollars in grid upgrades to address energization requests, but data center requests can be speculative, adding uncertainty to load forecasting and infrastructure needs.

In California, data centers can either help create downward pressure on rising electricity costs, as we just heard from PG&E, or they can drive up costs for everyday households and small businesses. Data centers will be more likely to help lower electricity costs if they're built, fully utilized, and operate over the long term.

When projects are speculative, when facilities use less energy than projected, or when they exit early, the cost of new infrastructure doesn't disappear. It shifts to existing customers. Because transmission costs are shared across customers, sustained and predictable usage is what allows those investments to spread more efficiently over time.

That's why managing risk is so important from a ratepayer's perspective. Protecting ratepayers requires a package of safeguards. There's no single tool that will solve this on its own. We've put together some guiding principles that provide a foundation for ratepayer protection. First, we should provide a continuity of revenues to ensure adequate cost recovery and discourage speculation.

Revenue continuity can be achieved by requiring deposits prior to establishing service that are returned only after a certain time frame or demand level is reached—early exit fees, minimum rate service term lengths, minimum demand charges, and letter of credit requirements. Second, we should prevent cost shifts by applying causal cost allocation.

Costs caused by new large data center loads should be allocated to the customer that causes the cost and avoid cost shifts to customers who did not cause the cost. Third, apply consistent regulatory practices across IOUs, including alignment of contract terms and cost allocation, to prevent preferential treatment in any one service area.

Consistent regulation promotes utility accountability to adhere to the same rules, it fosters competition based on an equal playing field, and ensures equitable cost allocation and protections for all California ratepayers. Fourth, ensure transparency and continued CPUC oversight.

Energization and rate design should include transparent processes with oversight from the Commission and stakeholder engagement and the ability to address any issues that arise in the future. The speculative and uncertain nature of data center requests for energization can inflate load forecasts and lead to overbuilding. We've already heard about some of this today.

It's critically important that the state's long term transmission planning accurately reflects the reality of data center development. Flexible transmission planning allows the state to reassess projects before construction if forecasted data center demand stalls or withdraws.

We provided comments in CAISO-led and CEC-led forums on the need to avoid speculative load forecasting and regularly reassess the need for new transmission projects, and we've initiated ongoing discussions with the CEC to ensure that we have the same understanding of these issues.

In this way, we can work together to help ensure that transmission investments are made only for projects that are certain to proceed. As data centers expand in California, the infrastructure costs required to serve them should not come at the expense of electric ratepayers.

By implementing practical cost responsibility rules that don't place undue risk on ratepayers and embracing flexible transmission planning, California can manage rising data center loads in a way that's fair and affordable for all ratepayers.

Happy to take any questions, and please make sure to visit our website, read our blog, and sign up for any public distribution of our comments and testimonies. Thank you.

Thank you. All right, turning back, Mr. Procos, too.

Thank you. I also have some slides. I don't know if they'll show up this time, but.

Second time's the charm.

Okay. There we go. Look at that. Sorry, jumped ahead there.

So, I'm going to start with a little bit of our resource stack—just go over this very quickly and you can see it's going from 2025 to 2035. Shows what we've been doing in terms of procurement to meet that increasing demand.

And the brown line up top is our forecasted load, and the RPS forecasted load is our compliance with the state renewable portfolio standard. We have a big commitment in the city to renewables. Being from the Bay Area, that's no surprise. We have a very robust portfolio of solar, wind, geothermal, some hydro.

We do have some natural gas power plants that we partner with the Northern California Power Agency on and we have 165-megawatt power plant in the city and the peaker as well. And I mentioned earlier how we're in the process of procuring 40-megawatts of batteries, which is currently under construction in the city.

The total in the pipeline, that's that kind of green kind of hashed line there, those are projects that we're working on right now. Anywhere from 5 to 10 projects at any given moment. We have staff that are dedicated to trying to procure this energy.

And I'm bringing this up because this is another challenge in this area, is trying to find the projects to be able to meet this demand. We haven't really talked much about resource adequacy today, but that's an important topic related to this. But it's something that's become increasingly challenging. At the end, I've got some slides that kind of layer on some additional challenges in this, in this area.

At the top boxes, those are the projects that we have inked. And so, they are in the various stages of being built. Whether they will all make it over the finish line is a different story.

Again, just want to emphasize the challenging time, but we are moving forward with that procurement. This slide shows the average system rate comparison with utilities in California, and we are proud to say that we are at the bottom.

We, we were number one at the bottom, but Turlock Irrigation District recently took over and so, we're working hard so that next year this will be back number one at the bottom there.

And we're going to have a separate hearing entitled, "What can we all learn from the city of Santa Clara and bring down utility rates in the state of California."

And when I term out, I'm moving out of PG&E territory.

You know, and I mean I think this speaks to the topic of the agenda that our partnership with data centers, I spoke earlier about how these are from a utility perspective, these are your best customers. I think their baseload and there's a lot of talk about the cost associated with them.

I don't think there's enough focus on the revenue that they bring in. There certainly are challenges in terms of are they going to show up and you need to deal with that. And it sounds like there's been some creative solutions that are being bought forward, and I'll talk a little bit more about how we deal with that.

But I mean a lot of the reason why we have data centers in Santa Clara is because the rates are so much lower. It's also the latency, so, but that's certainly a big benefit to our community. And then, turning to economic and rate benefits, developers directly fund dedicated substations.

So, when they're connecting on the 60KV system, and that's a large customer, typically 50 to 100 megawatts, they're paying for those substations, they're paying for those line extension. And we have something called a load development fee.

And this is a fee where we take into account all the projects that we're doing to expand our system and we allocate a portion of that into the load development fee. And that gets charged to new customers. Not all of it gets charged to new customers because our infrastructure, like many utilities, is older infrastructure.

And you have a new customer coming in a data center and they haven't been there, they haven't been using that infrastructure. It's not fair to allocate to them the cost of replacing aging infrastructure. It's fair to allocate to them the increased capacity that's driving that.

So, we try, we, we've tried to allocate it between those existing customers and the new customers. And just to give you an example of what that means, a 50-megawatt data center will pay under the low development fee for approximately $25 million, and that's not necessarily all upfront. It depends on what the agreement is.

It might be staged, but that's a significant contribution. The other thing that I want to mention is that we, like many utilities in the state, we've been building up our reserves over the last three to five years. And there's multiple reasons we're doing that. Wildfire concerns is one, one area.

The other issue is if you have a data center that, you know, we hear about the potential bubble and that may happen, but we're building up reserves to be able to mitigate that in the future. And the reality is for us is that 90% of our revenue comes from our large customers.

Right now, our residential customers pay 5% of our revenue. That number will probably drop as we increase. So, the folks who are funding that reserve are the large customers. So, we think that's a prudent way to approach that. I will say that I think somebody mentioned a letter of credit.

That's something that we are looking into as well, just in case to ensure that somebody, if we're spending money on these customers, that if they for some reason leave halfway through the game, that there'll be some sort of payment involved there. Developers pay—in our community, developers pay impact fees, so, they contribute to housing projects.

They provide high, stable, and predictable utility revenue streams, as I mentioned before. And from our perspective, that puts downward pressure on rates. Getting into discussions about are they increasing rates, it's a really complicated subject and it's a difficult one to kind of get.

I mean, I think that would take a whole 'nother day long hearing to talk about that. But in our model, we see it as a successful partnership, and it does put downward pressure on rates. And remember, when we charge those customers retail rates, embedded in those retail rates is the cost of transmission.

So, they are making a significant contribution to all that new transmission. They may not be covering all of it, but, and I do want to agree with PG&E that we do not think that specific customers should be directly allocated these kind of costs, data centers or whoever that might be. In California, transmission is socialized.

Everybody pays the same rate and building new transmission provides a lot of other benefits. It benefits other customers. It also gives the ISO and other operators, our own operators, the ability to operate the system more efficiently and have more tools in the toolbox. So, we are completely on board with the concept that was presented earlier.

Going back to the economic and rate benefits to our community, it increases property tax revenue. We transfer a significant amount of money to the General Fund and help with services in our community. And I touched on the e-reserves already and we really feel that the local, regional, and state benefits are significant.

I mean, I mentioned earlier, I mean Nvidia is based in our service territory. Intel is as well. And these really have done a lot in our area, not only in jobs but also financially. And then, I will close quickly on just some of the challenges ahead for us. Again, the policy uncertainty, this isn't really a new thing and specific to data centers.

Tariffs. I forgot to mention you have a handout that we provided that shows information on our service territory. There's also a one pager on—specifically on data centers. So, there's lots of more information than I provided here. And then, also, we're dealing with tariff issues as well.

As part of our $500 million system expansion, we've got a significant amount of orders which are overseas. And so, we're actively monitoring those. I know that's a bit of a moving target, but it adds some uncertainty to these projects.

Project delays, this relates to our procurement efforts and also, just building substations to be able to accommodate these customers. It's a difficult environment with supply chain tariffs and then other project delays. And then, finally, for US staffing, I mean, we've got a 22% vacancy rate. It's a tough environment to be to operate an electric utility.

I will say, I've got some of my staff here today. I mentioned earlier, we've got 235 staff members and they do an incredible job working on behalf of the community. I couldn't be more proud to lead this organization and have such a great workforce supporting us. And with that, I will close.

Thank you. And before we open it up for questions, turn back to Ms. Bunder.

Thank you and good afternoon, again. As mentioned, my name is Kara Bunder. I'm a State Policy Director at the Data Center Coalition. I just want to double click on a couple of comments that have been mentioned already today.

But at the outset of my comments want to underscore that D.C. members are committed to paying for their full cost of service for electricity and the infrastructure that they use. In addition to the, what my, my peer from PG&E was saying, it's not just limited to PG&E or utilities in California.

We're seeing this type of messaging coming from utilities from across the country. Georgia, which I had mentioned earlier, is a significantly and rapidly growing data center market.

Georgia Power, one of their main utility companies in the state, recently announced that they were able to have a three year rate freeze for all customers and also indicated that large load users, like data centers, are responsible for helping to keep those utility rates frozen for the next three years.

In addition to that, we always talk about Virginia again as the largest data center market. In late 2024, their Joint Legislative Audit and Review Commission issued a very comprehensive report looking at a variety of different challenges and opportunities surrounding data centers, ranging from cost allocations, to air quality, to noise, effectiveness of tax exemptions, etc.

But that report did find that, currently, costs are being appropriately allocated across all customer classes, including data centers, and that their State Corporation Commission, the equivalent here being the PUC, is appropriately positioned to address any future cost allocations through their expertise and processes that are in place.

I have other examples from other utilities across the country that I'd be happy to share, but in the interest of time, I'll move on from that.

But additionally, so we talk about, we also talked about SB 57, from last year, today, in the interest of transparency and of course many of you know, as I've spoken on committees and provided testimony on that Bill, DCC was opposed at the, at the outset of that Bill because for two primary reasons.

One, the singling out of data centers as one end user of electricity and two, because we do agree with the premise that the CPUC is the appropriate authority to be able to address those cost allocations through evidence-based deliberations and transparent hearings. To that end, which is also like the flag, that Data Center Coalition primarily, again, through our energy team, has been engaged with a number of those regulatory proceedings across the country.

And as part of that, we have supported the traditional rate making principles including transparency, nondiscrimination of an end user of electricity cost because cost causation, and also of course, limiting that cross subsidization that would impact other ratepayers in, you know, covering costs that are being attributed to data center growth.

So, we encourage utility commissioner—commissions—to keep rate design and cost allocation transparent and evidence-based to allow independent experts to clearly review how costs are assigned and test those methods.

And again, double clicking on some other things that were discussed before, as part of those regulatory proceedings, you know, we've mentioned, we've mentioned minimum demand charges, minimum contract terms, exit fees, other ways of ensuring that they are not stranded assets and ensuring that those costs are not passed on to other customers.

So, as we continue to engage in those regulatory proceedings across the country, we're seeing different types of settlements that are coming to terms but would also be able to follow up with information on where DCC has been involved and what that's been looking like.

Thank you. All right, questions from Committee? Mr. Rogers.

Thank you, Chair, and again, thank you so much. This has been a really fascinating discussion, and I've heard a lot about if we do this well, it can have a positive impact on ratepayers. If we don't do it well, it could increase their costs.

But I want to acknowledge that what I haven't really heard is that it's a policy choice on whether it impacts them at all.

And last year, I worked on a bill to do green hydrogen production that required changing or amending the over the fence rules and had a surprising amount of businesses actually reach out asking us to apply that to the data center conversation as well.

So, I just kind of want to ask. We're talking and I understand why the IOUs and the utilities would want the added load, especially when we talk about rate of return around infrastructure and markup on energy.

But why are we not also talking about whether or not to open up a scheme where in which high energy usage, like a data center, can co-locate above the 50-megawatt limit to be able to produce their own energy and drive that conversation as well?

I'm happy to start there. Great question. And I think actually Chair Petrie-Norris made a reference to this. So, in a situation where a data center is co locating a generation resource, there are a couple of things happening there.

So, by them leveraging that co-located resources, they are not contributing to the fixed costs of the system that help serve them. So, I think a big example of that is the wildfire mitigation costs that we see in the state of California that support customers all over the state.

And so, the more that they co-locate that resource, the less that they help support those fixed costs.

Also, in that scheme, that customer typically wants—obviously, typically, like 99% of the ones that I talk to would like to have full access to the grid as a backup should that co-located generation go away, has to go down for testing, or malfunctions, or something like that, which means the utility, whether it's PG& E or Silicon Valley Power, is still making those capital upgrades to the system to that full extent to be able to serve them if they have to switch over to the grid.

And so, those upgrades are paid back to the utility and paid back to other ratepayers by charging that customer.

And so, if that customer is not paying a minimum demand charge or is not pulling electricity from the system, it's more challenging for the utilities to be able to recover those revenues and for us to have this concept of spreading out these fixed costs across a larger number of kilowatt hours.

I'm sure the utilities have comments on that as well.

Yeah, and I'm not saying I'm advocating for it. I'm asking about thinking outside the box on this as we're talking about wholesale changes that we do need to make. We're having the same conversation about electric vehicles and the infrastructure that they use. And the scheme that's out there currently is a fixed fee that they pay into the pot.

You could have the same scheme for fire prevention for all of the other assets that you're talking about that currently we pay through ratepayer funds, which we're trying to move away from in some instances. There are other ways that you could get at the same thing and have responsible payers still help fund that infrastructure.

I just want to comment on that. So, the Public Utilities Commission, about two years ago, did adopt what we're now calling the base services charge. It was called an income graduated fixed charge. This was in response to legislation that required us to do so. And so, we have set that infrastructure charge at about $24—between $24 and $25—for the investor-owned utilities.

This is something that—I can't speak for Silicon Valley Power, but other public power entities like SMUD have had on their bills for decades. And so, that was something where we were very far behind on the investor-owned utilities.

And so, we have done that at that $24 and I think 15 cent level. I will say that there has been request from stakeholders for us to keep it at that level for a point of time to study it more closely, see how it's matching cost causation principles.

And so, we, you know, may look at that fee in the future. But I think you're making a great point. It's just that right now from stakeholders, there's been an ask for us to see how it pans out. And so, just an update there.

San Diego Gas and Electric in Southern California Edison did their implementation of that base services charge or infrastructure charge in Q4 of 2025, and it's being applied in Q1 of this year by PG&E Liberty Utilities in Bear Valley.

And I mean, just to add to that, I mean, we're, you know, we're not, obviously we're not regulated by the CPUC. So, our local regulatory agency, typically our city council, makes decisions on this. We don't have a base charge. But couldn't agree more with the comments made by the CPUC.

There's a lot of sensitivities and for us, you know, getting these conversations about standby charges and you have a customer and it kind of defeats the purpose.

Really, the way I see it is that they're supposed to be providing this revenue, rebuilding infrastructure to accommodate them and if their generator goes down, then they expect us to back them up and that causes some heartburn. So, I think, you know, we look at different ways of kind of how do we approach that.

And I will say that we have a lot of interest in our jurisdiction to—from data centers—to partner with generation and whether that be a fuel cell or something of that nature and that's something that is—we do have some applications of that. It's very limited.

But I think there is a lot of concern about the standby issue and how would we then how would they make their fair contribution towards some of those costs?

Understood. Another question, and my last question. Which states do you think are currently doing this well or which states do you look at as sort of a model, whether for a good reason or a bad reason, as you have these discussions and especially as you start kind of down the path on how to best implement?

I guess the answer depends on if you're talking about being able to interconnect customers quickly. There's a lot of states that are, you know, doing things from a permitting perspective, etc.

I'll clarify, just in terms of protecting ratepayers.

Yeah, I guess when I look at the news...

That's from increased costs.

When I look at the news, I don't—doesn't seem like many states are. It seems like many states have just taken the traditional approach.

All these new costs, we're going to allocate them to all customers and that's why you're seeing customers rates across the board go up 5, 6, 8, 9, 10%, whatever it may be. We have the opportunity here to do something different and that's what we're talking about.

And I think a lot of people propose the things that that would make a lot of sense and help us avoid that situation in California.

I think my concern is that it's very difficult to make apples to apples comparisons with other jurisdictions. I mean we are unique. They're unique.

I mean there was a comment earlier about some of the challenges in PJM with the capacity and some of those costs and I think it's very easy to see headlines and maybe draw some conclusions. But you know, for us, I think it's just a different situation.

So, I get a little nervous when people start thinking about how should we be doing it, like Virginia, or what's the lesson learned? I mean, I think, I mean, you asked which state is doing it really well. We're not really a state, we're a city, but I feel like Santa Clara is doing a pretty darn good job there, so.

Why didn't you put that in your presentation at all then?

And there are other states. They're not just doing the status quo. Some of the risk mitigation tools I mentioned earlier are being implemented in other states. So, they are very much looking at minimum demand charges, contract terms, ensuring that they have payments and that data centers are paying their fair share to take that risk off of ratepayers.

Ohio is one of them. Virginia is looking at different tools. Basically, any of the states that already have quite a number of data centers are grappling with these same issues, and they are adopting risk mitigation measures. Some of them are quite extensive and further than we've been talking about here in California.

I'm happy to make just a comment there as well as it's not a one size fits all. And I also think about California being unique in the clean energy and greenhouse gas reduction goals that we have created for our state.

So I think Texas was brought up as an example of a state that has a lot of data center growth. A lot of those data centers are being brought on the system by co locating with natural gas facilities. That's the way the vast majority of them are connecting to the system.

And that's not really in alignment with our clean energy goals. So we would like to see as much as possible data centers getting connected to a clean electricity grid that continues to grow more clean and stays reliable while protecting ratepayers.

All right, Assemblymember Erwin and I was just going to start my conversation that way that it is that it is critical that we maintain our clean energy goals. But I think for all of us, all we hear about is how high energy costs are.

So we need to do everything we can to make sure that these costs of adding data centers are not being passed on to consumers. And certainly I know there's some. It sounds like from the data center coalition that there's cases where you can really prove the savings because you have a much larger load.

But PG&E has talked a lot about the idea of being able to reduce the average household Bill by 1% for every gigawatt that has been added. But the entire day today we've been talking about all the uncertainty and is the load going to show up?

And so how certain are we that households over a 10 year period are really going to see for every gigawatt a 1% reduction?

That's, that's right. In Rule 30 we provided a forecast and based on our assumptions, the number of assumptions that if these things occur, this is a savings that we think could be achieved for the average residential customer over that 10 year period.

The big, the big item that we see that needs to be addressed, we rule 30 was just about the interconnection cost, right. To connect to the grid.

We think the big item that needs to be solved, one of the big items needs to be solved is what about all the incremental generation supply ra that's going to be needed to serve those data centers. We need to get that cost allocation, that rate design correctly.

I think the PUC is going to open a large load proceeding here shortly and that's one of the biggest issues that we would like to see solved there. Because we think if we do it right there that supplements the analysis that we did and will drive those savings.

Okay, and then for the original Rule 30 proposal, was there really a guarantee that the cost wouldn't be passed on? It seems like there was a little disagreement between PGE and the Public Advocate's office on the original Rule 30 proposal.

Go ahead.

Yeah, I'm happy to answer that. We did provide testimony on the Rule 30 proposal. I think the original proposal did not include things like minimum demand charges or set contract lengths or early termination fees. We're very happy that PG and E has amended its contract, its proposed tariff.

We still think it should go further so that proceeding is still, you know, in action. I'd also just point out that regarding PGE's claim about the for every gigawatt a 1 to 2% rate reduction. We're doing our own analysis on that. We'd be very interested to see PGE's analysis to how they got to those numbers.

We did look at their 10 year horizon analysis in the Rule 30 proceeding and we thought the assumptions that, that you mentioned were extremely unrealistic.

We did our own analysis just as a bookend to show that if data centers, rather than serving a 10 year term during that period, if they exited it at year three, it would actually add over $6 to a monthly Bill for an average residential customer. So the risks are real and it very much depends on the assumptions.

Yeah, I will say we did have an original Rule 30 proposal it was based on. It wasn't just PG and E coming up with. We did a lot of outreach too, external stakeholders, including our regulatory bodies, customers, etc. To design the original rate 30 through the proceeding.

There were other criteria that were that were offered up that we thought that makes a lot of sense. So we amended our proposal to include the minimum demand charge, early termination fees, a contract term to make sure we had added customer protection.

So more than willing to consider things that are going to again get to the point of not having having our existing customers be exposed to higher cost.

Right, thank you. And then I just had one more question for Kara. We talked a little bit about PJM which by the way, the J is Jersey. I had Pennsylvania, Jersey and Maryland. They are the ratepayers in that area are seeing tremendous costs increase.

But can you talk a little bit more specifically on what we have to do in California to make sure that we are not repeating those mistakes?

Yeah, I think some of this has been covered earlier and that this is not really an apples to apples comparison of PJM to CAISO to, you know, ercot, we can't really compare these different markets and what's happening there necessarily.

But I think some of the items that we already talked about today and being able to do these regulatory proceedings for different tariffs or cost allocations across all customers with some of the items that were mentioned before, the early exit fees, the minimum demand charges, the, just the minimum contract terms in and of themselves can help to ensure that those costs are spread across customers appropriately, but also addresses that risk of stranded costs to get ahead of that.

Again, I mentioned SB57 earlier and how we were originally opposed to that, but we do think those regulatory proceedings should be based on low characteristics and load profiles and not on the end user itself. You know, I think this is something that PJM is going to be grappling with as well.

More broadly, it's certainly an issue, but I wouldn't say that there's a similar operation undergo being undergone here as there is across those 13 states that PJM covers.

And then do you, so you see, I know a lot, a lot of it has been covered over these few hours, but do you see these, the, the proceedings going in the right direction then? Are you in agreement with the public advocates and some of the, the, the guardrails that they would like to see?

So we've certainly reached settlements in several of these regulatory proceedings based on what those explicit terms have been. There have been others that we're contesting just based on what the, you know, the specifics for what those minimum demand charges are, what those minimum terms are and for those specific jurisdictions.

But again, we're coming at this from the evidence that we're bringing to these proceedings and how that ends up shaking out in those regulatory proceedings. So again, we've had settlements and in other places we're contesting the settlements just based on what those final terms end up looking like.

Can I make a comment in response as well, not in disagreement, but just to say that there are fundamental differences between the PJM market, which is an electricity market, which is different from the conversation we're having about cost allocation and how rules are, how the rules are created, the customers in PJM that are facing these skyrocketing costs.

That's occurring because the PJM has been set up as a capacity market where the increase in resources on the system is driven by price signals that encourage generators to build new generation. And so they have been struggling with those price signals and new generation has not been built.

The California Legislature adopted Senate Bill 350 which created integrated Resources Planning, which leverages the expertise of the Energy Commission, Public Utilities Commission and the California Independent System Operator to have resource procurement and build out regime that is based on the reduction of greenhouse gas emissions.

And that occurs because the Public Utilities Commission orders load serving entities to enter into contracts and build new generation. So you're not just depending on a capacity market like PJM has to create high price signals to get generators to build new generation. So through the existing state mandates that are in place, we have been driving new generation.

That was my slide. And so that's how we're building new generation. It's not through these high price signals the way PJM is set up. It's just a fundamentally different regime that California has adopted in order to protect ourselves and continue to meet our greenhouse gas reduction goals and reliability and customer affordability.

Thank you.

And I'm going to jump in because I have a couple of follow up questions related to Rule 30 specifically and then we'll turn to Assemblymember Papen and Assemblymember Zabur.

So I think one of the reasons that Mr. Prokos, we invited you and wanted to hear from Silicon Valley Power is that in many ways you're the tip of the spear. And, and while everyone's guarding against comparing jurisdictions, I think that California and the rest of California very much needs to learn from the experience that you've had.

Because if I understand correctly, you have the highest density of data centers in the state, you have the second to lowest, about to be the lowest utility rates in the state. So in some way it feels like we're overcomplicating this. And I would love the PUC to respond to why.

Why are we not just implementing a rate structure that models what Silicon Valley Power has done successfully?

So I'm not here to advocate on PGE's behalf their application. That's an open proceeding. That's before the Public Utilities Commission. So we are considering the proposal they have put forward, but we're also considering the perspectives of stakeholders like the public Advocates office and other rate payer advocates and many other different type of stakeholders.

So yeah, I'm not here to advocate on what they have put forward. And it's an open proceeding before the Commission. So we will be taking all of those things into consideration. How do we protect ratepayers?

How do we make sure these new loads get online and how do we get the cost allocation right to make sure that we are balancing those interests with the focus on the ratepayer benefit? If I misunderstood your question, let me know.

But I just hopefully didn't make it seem like I was advocating for the regime that PGE has put forward. It's under consideration by the Commission. The interim tariff that we adopted last year did not include the cost allocation piece.

Right, Understood. So then maybe I'll turn it to PG&E for, for your proposal. And I, you know, I, I don't not like your framework, but if anything, it seems like SBP actually includes more costs upfront. If I understand kind of how the. Yes, the developers directly find dedicated substations, line extension, low development fees, supporting capital projects.

Well, I mean, I think, well, I'm not sure exactly what goes into their, what their fees are, but we do.

Customers are paying for our proposal and how we've been implementing as customers pay for what we call the type 12 and 3 cost to be able to connect to the grid, the line extension upgrades to the substation, other work to be able to connect to the grid.

So I think we're, I think we're doing what they're proposing. I'm not sure I would have to look deeper in it to see how it differs.

I mean, the only thing I would add is the low development fee. Is that where we take into account the amount of money that we have to spend on the system. That's that 450 to 550 million that, that we're undertaking to go to from 740 to 1300 and we allocate a portion of that to those new customers.

And I gave that example where a 50 megawatt data center could potentially pay up to $25 million. That's separate from the substation cost. They pay that and they pay the line extension. So it is, and I will say that it is.

I'm not aware of any other fee of that nature in the state, if maybe others in the country. I don't know. There's different forms of that, but I mean we've had that for a little while.

It's worked really well for us and it covers kind of strikes that balance in terms of ensuring that you're kind of fair, I hate to use the term fairly, but you're fairly allocating those costs between those existing customers and the new customers.

Okay, and then another question, Mr. Madero. So in your proposal, the customer would pre Fund the loan and in your proposal, how would the rate of return calculation shift as a result of that pre funding?

When you mean the rate of return, are you talking our rate of return or the rate of return, the refunds that we would provide to customers?

Your rate of return on the Project. So.

Right. So customers pre Fund that work. Yes. And. Right. And we measure their revenues over the course of time. They have the opportunity, not the guarantee, to be able to earn refunds if their load and the associated revenues materialize.

Materialize over time. So we look at that on an annual basis. If it does materialize, we provide them a refund for that amount. It's at that point we have to then borrow the money. Right. To be able to pay the refund.