Hart Energy Q&A with Ravi Srivastava: AI’s Gas Demand “is Going to be Huge”
The following is a transcribed Q&A with CNX’s Ravi Srivastava (President, New Technologies) at Hart Energy’s annual DUG Appalachia Conference in Pittsburgh, PA.
November 27, 2024
Jordan Blum, Hart Energy: I just wanted to start things off getting your opinion on just really how big of a deal [artificial intelligence] is going to be for domestic power demand in terms of actual increases. Most of the talk's about data centers and AI, but it's obviously more than that as well.
Ravi Srivastava: Yeah, so first of all, I think AI data centers, these are opportunities of a lifetime, opportunities, like, generational opportunities here. So, I think every bit of focus, every bit of rigor that we can put behind to make sure these industries, these opportunities are successful, we have to do that. And at the core of that, to make any data center work, to make the AI technology work, what they need is power. So, I think rightly so, to make the technology commercial, to make the technology better, there's a humongous demand for power.
I think every day I get an email with an updated forecast for, 'Hey, our last forecast was wrong. We're going to need even more power than what we said in our last forecast.' So not a day goes by where these forecasts are [not] getting adjusted. And at the same time, these AI data center demands [are] going up and the policy is pushing for electrification of everything else, whether it's transportation, whether it's buildings, whether it's, you name it.
And all that stuff is putting a lot of pressure on the grid. And we see natural gas as the natural solution to be able to address that demand in the short term, in the long term. I think it's a tried and tested, proven way of addressing this demand. Whether the pace is slow, fast, whatever it is. I think the operators here, like CNX, we stand ready to be able to answer that call.
How challenging is it, or not, to convince technology companies and everybody building these projects that natural gas is lower emission? It is not just reliable and affordable, but a sustainable approach?
Ravi Srivastava: I mean, that's a great question. So again, if you look at the last 20 years and the energy transition that's taken place, and natural gas is playing a much bigger role in all forms of energy utilization, whether it's in some form of transportation, power generation, industrial uses. And that single handedly has offset more CO2 emissions from U.S. atmosphere or global atmosphere than a lot of other technologies combined, right?
So natural gas is a very well-established solution from a low carbon intensity standpoint. And then on top of that, this particular region, we're passionate about Appalachia. We're passionate about opportunities here – Marcellus and Utica Shale. A number of studies have shown that the lowest carbon intensity or emission intensity gas is produced in this region. And then you take it a step further where we're not stopping at that. We're not sitting idle that, yeah, this basin is great and we're going to stop doing anything from this point onwards.
The focus from CNX and a lot of other operators is to just keep driving that emission intensity lower, and a lot of investments are being made into technology and solutions. CNX is developing proprietary technology to decarbonize this value chain, and we've been able to get into numerous JV's and commercialize some of these solutions to not only decarbonize that value chain for ourselves, but for our peers.
And then we're also proud of our waste gas capture program where this is waste methane that would've gone into the atmosphere. So that is also a solution that we provide as a feedstock for this power generation. So, there are numerous ways to address the sustainability needs that a data center outfit might have for a low carbon solution for the power generation. And from a resiliency standpoint, I just can't see any other solution that even comes close in terms of being able to provide a resilient feedstock that can deliver power with the reliability that natural gas can.
Very good. You touched on it some, but can you elaborate a bit on just some of the technologies you're most excited about in terms of lowering emissions?
Ravi Srivastava: So we have developed this proprietary technology and we've gotten to a JV with an oilfield services company here called Deep Well Services and our AutoSep technology, what that allows us to do is take emissions. It's a closed loop, high pressure flowback solution where a lot of emissions, a lot of methane goes into the atmosphere during some of these completions and flowback processes. But our solution is a closed loop process that allows us to conduct all this stuff, all the routine parts of this operation [while] avoiding any methane going into the atmosphere. So, there is emission reduction that's happening by virtue of using this technology.
And then the technology also allows us to harness this natural gas and separate solids and liquids from a high-pressure gas stream. So, energy that's typically wasted in an oil and gas process, we are able to harness that energy and convert that into power. So, to address other power needs that we might have on the location or do other things with it. So, all that helps reduce the emission intensity of the operations further. That's just one example. We also have ZeroHP CNG technology, micro-LNG technology that we have developed that uses the technology at its core. I could talk at length. Depends on how much time we have.
What’s the view on data center electricity demand growth? What role does the natural gas industry play in addressing that demand?
Ravi Srivastava: So, I think, like all of us in the natural gas industry over the last year or so, not a day goes by when I don’t get an email that updates what the data center demand’s going to be. I think it’s going up… the projections of the forecast are going up on a routine basis.
Every year, [PJM’s demand] forecast just keeps going up and data centers is just one part of this. I think policy is driving electrification, not just on the AI needs or the power generation needs side of things, but also transportation and buildings. There’s a lot of stress on the grid from all these different types of needs.
Then secondly, there’s a lot of need for policies driving the need for lower carbon solutions for this power generation. So, a lot of renewables [and] nuclear is in the mix. The forecast I see is, I think Toby [Rice] said 10-18 Bcf is what may be needed to address the need, the demand growth that we see on the power generation side of things. There are also forecasts out there that say natural gas is supposed to be only 50-60%. The rest of it is supposed to come from renewables and nuclear. But if you look at the history of what’s happened in that space, I was reading about this solar project that requested a PJM interconnection queue in 2018 and it was in April 2024 that it finally went online. So that’s the amount of time it’s taking for renewables to address some of these demand needs. Over the last 30 years, there’s only been one nuclear power project that went online and it was more than 200% over budget and about a decade behind schedule.
The demand is going to be huge, the 10-18 bcf number that’s out there. Folks talk about solar, wind, [and] nuclear playing a role, but the proof is in the pudding where we haven’t really seen the response from other sectors to be able to address those needs, so I feel very good about the role that natural gas is going to play in addressing those needs. I think natural gas provides a very resilient and a clean solution to address, not just the data center needs, but as we move toward more electrification of other industrial needs, natural gas is going to have a bright future ahead.
Is there a case to be made that maybe onsite generation could be even more reliable in some ways?
Ravi Srivastava: There’s a lot that we already do in the industry where we have to provide a certain level of reliability in the operation of a lot of our facilities, so for us to be able to deliver natural gas to a certain location, especially onsite power generation location – I feel very confident that the work this industry has done – we’ve talked about all the accomplishments that folks here have achieved over the last 15, 20 years.
To be able to deliver those molecules at a location, it’s not going to be a problem. I think the industry stays tall to be able to address those needs. In the data center world, they talk above the ‘five nines’, the 99.999% reliability of whether the electrons [are] going to be there.
I’m going to make a bold claim here – that if we do some of these onsite generation type solutions, especially in this Appalachian Basin where the operators in this industry operate and deliver the molecules – I think that’s an easy number to beat, to be able to deliver, and the solution that AMP that they can provide with the equipment just being there, to me, there’s argument to be made that it’s a more reliable solution than relying on the grid.
We’ve seen what’s happening with the grid and some of the reliability concerns that folks have because of the intermittency of other capacity that’s been added to the grid.
Are they telling all of you, ‘We need more natural gas. We need producers to produce more’? And how do you balance that with what we’ve heard a lot today about growing demand for LNG – so you now have a couple of different avenues asking for more natural gas? But what are data centers saying in terms of increasing production?
Ravi Srivastava: I think data centers are looking for electrons, right? I think they want a reliable and a large supply of power is what they’re looking for. We’ve been talking to different data center outfits. I can tell you that about a year and a half ago, the conversation was a lot different. I think at that point and time, the demand for an electron coming from say, fossil-fuel-based power, including natural gas, was not as high. I think they were looking for a cleaner form of electron but over the last year and a half or so, I think that that table’s flipped. I think at this point in time, what I understand is that speed is the immediate need – how soon can I get power for the capital investments that they’ve already made.
The second part of it may be around, then how do you maybe clean it down the road or further? But all the capital investments that’s been made right now, the need is to get the power, but the reliance on grid is extremely unpredictable and unreliable.
I think there are folks that are waiting in line for a lot of years to be able to get that. Part of me feels like what other solution is there except for off-the-grid, maybe behind the meter, you know, like a grid solution where we can get natural gas with, there’s very effective modern turbine power generation technology that are available to convert those gas molecules into electrons and then folks like Heinz, they come in and they provide the backbone that’s needed for the data center to operate and into what they need to do. I do feel like the building blocks of – once people get a better perception of the grid solution – is not going to be there unless they want to wait six years, then, I don’t know, maybe the AI revolution would have passed by at that point in time. But if they want the solution now in the next year, two years, three years, this is a solution.
I think the markets are very efficient. These hyperscalers and data center companies, they’re making big capital investments. There’s a reason why Nvidia stock is going through the roof – because people are buying hundreds of millions of dollars’ worth of these chips and they don’t want them to just sit idle and once they’ve made those investments; they have to figure out a way to run them.
And if the grids are not going to be able to provide that solution in a cost-effective manner, in a reliable manner, that creates an opportunity for a different sector of the market to bridge that gap, address that need and I think that’s where we come in and there’s an opportunity for us, especially this region. We have a lot of ingredients that are required to make a solution like that work.
And maybe you’ll see some solutions like that pop up in the near term to kind of set the example for bigger projects to follow.
Turning to a little bit toward the sustainability question of this is, how can the industry address the resiliency requirements for data centers? And also, many big data center players who are looking for zero carbon, or clean energy solutions, what can the industry and specifically CNX do to address this?
Ravi Srivastava: Again, I think I mentioned earlier, policy is kind of driving the markets, the industry toward zero carbon, low carbon renewable solutions, but that does introduce a lot of risk. I mean, we’re seeing that in the grid. I don’t know if people are familiar with this ELCC curve that the grids kind of – it’s a metric they maintain to kind of measure how reliable the grid supply is going to be. We’ve been tracking that for PJM, and it’s been coming down continuously. It was at 51% a couple years ago. It’s dropping and they’re saying by ~2030 timeframe, it’s going to be 28%, it’s going to get cut almost in half. What that means is, with the influx of a lot of renewable sources, solar, wind, which are more intermittent in nature, the grid’s ability to deliver whatever the peak power is, whatever the demand is, it goes, but that’s where solutions like what we’re talking about, or even like natural gas solutions in general are just more resilient.
All these grids publish what the ELCC ratings of different solutions are, and natural gas is right up there. If you’re looking for a resilient solution, you’re not going to find a solution that’s better than natural gas-based power. And again, this region is blessed with abundant natural resource and at the same time through the efforts of the work that folks in the room and our industry have done...
...this Appalachian Marcellus and Utica reservoir is one of the cleanest in the world, so we have a very, very solid foundation to start off with where an electron produced or power produced from a Marcellus or a Utica gas is, by nature, it’s going to have a very low carbon intensity to begin with.
I know a lot of people sitting in this room, we’re not going to sit idle at that point ever, as a basin were producing low carbon intensity, we’re going to stop there.
CNX has already been there for a while where we’ve been generating these offsets and environmental attributes for a long time. And there’s other things that the industry’s doing. I think there’s a lot of work happening in the – what I would say – the upstream and midstream value chain where folks are trying to reduce emissions further, whether it’s on the pneumatic devices... We have developed some proprietary technology ourselves – the AutoSep and the Zero HPCNG and the Micro LNG solutions – these are all solutions that help reduce the emissions that are associated with that upstream and midstream value chain. So, you can drive that emissions intensity down further and the second part of it is downstream of this, when power is being produced, carbon capture and sequestration – it’s a pretty well-developed technology.
It’s happening in different parts of the U.S.; it’s happening in different parts of the world. We have to identify it within this region where the opportunities are to sequester that CO2, but that’s another pathway to get to a low carbon solution for power generation.
Lastly, one thing, something that we are extremely proud of at CNX is our methane capture program where we capture waste methane that goes into the atmosphere and use that methane for beneficial use[s] such as power generation. If you look at the lifecycle assessment of methane going into the atmosphere and by virtue of capturing it, that’s a significantly carbon-reducing action. There are numerous established pathways that – whether it's in LCFS, California Air Resources Board – there are different programs that value the act of capturing waste methane from the atmosphere. If a billion cubic feet of methane does go into the atmosphere, that’s almost a half a million tons of CO2 that’s liberated into the atmosphere, right? It’s probably more than that, but by virtue of capturing that, you have a negative carbon intensity product that’s available to produce natural gas that can then be used for power generation and other sources. There’s no reason why you can’t use a blend like that – of natural gas and this waste methane to have a carbon neutral blend available for data centers or even other sources.
Recently, Microsoft announced a deal with the Three Mile Island – 800 megawatts or something over the next 24 years for zero carbon power. There’s no reason why a similar transaction can’t be done with the solutions that we have in place as well.
What attributes does the Appalachian region bring for development of data center infrastructure?
Ravi Srivastava: I want to start off by saying that I’m kind of surprised and mystified [by] why large-scale data center infrastructure development has not taken place in this region. From what I understand, what’s required, what the key ingredients to make this happen [are] low-cost energy – check – we have that. Locations that are maybe close to metropolitan areas, so you have access to talent pool like markets and all that stuff – check – we have that. Universities for talent pool – we’ve got CMU, Pitt, like really good talent pool – we have that. It sits on a fiber network, from what I understand, where the opportunity to connect into the fiber grid and all that stuff is going to be there – we have that.
So, I think it’s only a matter of time. I think we bring a lot of key ingredients required to make these things happen and maybe this variance in supply and demand for the electrons is maybe what’s going to ultimately shine the light and bring those opportunities to the region, but we’re excited about the potential of Appalachia to service that demand.
You Might Also Like:
