From Gas Engines to Data Centers: How Distributed Energy Found Its Moment
Reflections on 2008-2024
Clarke Energy in 2008 was not what it would become. It was a Liverpool-based business with deep technical roots in gas engine technology, built around the Jenbacher engine range and a clear focus on delivering distributed power solutions to industrial and commercial customers. The company had been operating since the 1980s, had genuine engineering depth, and was beginning to think carefully about how its core capability, deploying and maintaining reliable onsite generation, could extend into adjacent markets. One of those adjacencies was biogas and anaerobic digestion, which is where I came in.
My initial role was to support the marketing and commercial development of the Haase MBT-AD system, a German mechanical biological treatment and anaerobic digestion technology Clarke Energy was looking to establish in the UK market. It felt, at the time, like a continuation of the work I had been doing; the same fundamental question of how to take a technically credible system and build the commercial and regulatory conditions for it to succeed. The waste sector was familiar territory. The technology was different but the challenge was recognisable.
What I did not fully anticipate was how quickly the work would broaden.
The engine at the centre of everything
The Jenbacher gas engine is a remarkable piece of equipment. It is designed to run on a wide range of gaseous fuels, natural gas, biogas, landfill gas, sewage gas, coal mine methane, hydrogen blends, with high efficiency and very high availability. That fuel flexibility is not incidental. It is the engineering principle that connects almost everything Clarke Energy does across its markets. An engine that can run on landfill gas today can run on biogas tomorrow and on a renewable gas blend the year after. The asset does not change. The fuel pathway does.
That insight took time to fully appreciate, but once understood it reframes what distributed energy is actually about. It is not about any single fuel or any single technology. It is about placing reliable, controllable, fuel-flexible generation capacity at the point of need and designing systems that can evolve as the energy landscape around them changes. This is the principle that would eventually connect waste-to-energy, biogas, CHP, microgrids, and data center power into a single coherent body of work, even though at the time those connections were not yet obvious.
Scaling internationally
By 2011, my role had expanded to cover Clarke Energy's international marketing, then across ten countries. That breadth introduced a set of observations that would prove increasingly relevant over the following decade. In markets where grid infrastructure was constrained, unreliable, or simply absent, distributed generation was not a supplement to central power; it was the infrastructure. Extractive operations in remote Australia, industrial facilities in sub-Saharan Africa, agricultural processing plants in developing economies: these were environments where the questions about reliability, fuel supply, maintenance logistics, and operational continuity were not theoretical. They were immediate and the consequences of getting them wrong were serious.
Those environments produced a practical education in what resilience actually means as an engineering outcome. It means thinking about fuel storage and supply chain vulnerability. It means designing for maintainability under difficult conditions rather than optimal conditions. It means understanding how a system behaves when something unexpected happens, not just when everything works as planned. And it means designing with the full operational life of the asset in mind rather than just its commissioning performance.
This thinking would take years to become fully articulate, but it was forming in those early international projects in ways I did not entirely recognize at the time.
A conversation that had already started
By 2016, something was shifting in the data center sector that those of us in distributed energy had been watching for some time. Facilities were growing. Power densities were increasing. The questions being asked at conferences were beginning to include onsite generation, CHP efficiency, and the long-term adequacy of grid supply for large-scale digital infrastructure. I spoke at DataCentres North that year on exactly this topic not as a prediction but as an observation of a trend that already seemed underway.
The conversation was early. Most of the industry was still focused on standby diesel resilience, renewable energy procurement, and PUE optimization. But the underlying infrastructure question; how do you supply reliable, large-scale power to facilities whose operational requirements are continuous and unforgiving, was one that distributed energy had been answering in other sectors for decades.
In 2017, I was part of the UK government pavilion at the United Nations COP23 in Bonn, speaking on hybrid power generation, microgrids, and grid-balancing technology. The framing there was around energy transition and smart systems flexibility rather than data centers specifically, but the engineering principles were identical. Distributed, controllable, fuel-flexible generation. Systems designed to interact with wider networks rather than operate in isolation. Infrastructure that could evolve rather than be replaced.
The acceleration
The period from 2019 onward changed the pace of everything. Clarke Energy's acquisition by Kohler, and subsequently the formation of Rehlko as a standalone energy business, brought new resources and a sharper strategic focus. The decision to relocate to the United States in early 2022 reflected where the market was moving fastest: a country with extraordinary data center growth, significant grid constraints in key markets, and an increasingly urgent conversation about how AI infrastructure was going to be powered.
The AI data center conversation that became industry-defining in 2023 and 2024 was not, from where I was standing, a sudden development. It was the arrival of a question that the distributed energy sector had been positioned to answer for years, about what happens when grid timelines extend beyond investment horizons, when power density exceeds what traditional infrastructure was designed for, and when the cost of downtime becomes large enough that reliability is no longer a specification but a strategic priority.
The election as Vice President of the COGEN World Coalition in 2024 was, in one sense, a recognition of that moment — the point at which CHP and distributed generation moved from a background conversation in the data center sector to a central one. But from another perspective it simply reflected a body of work that had been building since the early Clarke Energy years, through the international projects, the Africa engagements, the microgrid conferences, and the gradual accumulation of evidence about how distributed energy systems actually behave under real operating conditions.
What the arc looks like in retrospect
I have been asked, more than once, whether the move toward data center power represents a pivot, a decision to reposition around a fashionable topic at the right moment. The honest answer is that it does not feel that way from the inside. The questions I am working on now; how do you deploy reliable power quickly in a constrained environment, how do you design systems that can evolve as fuel pathways, carbon constraints, and grid conditions change, how do you balance speed and long-term performance in infrastructure decisions that will last twenty years, are recognizably the same questions I was working on in different contexts from the beginning.
The waste sector taught me that technology readiness and market readiness are not the same thing. The international projects taught me that resilience is a systems outcome rather than a component specification. The biogas and CHP years taught me that fuel flexibility is an asset that compounds over time. The microgrid work taught me that the control layer is where the real complexity lives. And the current data center moment is teaching me that when all of those lessons converge on a single sector at a single point in time, the opportunity to contribute something genuinely useful is significant.
That is what the last two decades of distributed energy work has been building toward. Not a pivot. A preparation.
This is the richest article of the three — it covers the full arc of the career. FAQs grounded in the content:
Frequently asked questions
What is the Jenbacher gas engine and why is fuel flexibility significant?
The Jenbacher gas engine is designed to run on a wide range of gaseous fuels including natural gas, biogas, landfill gas, sewage gas, coal mine methane, and hydrogen blends, with high efficiency and very high availability. The fuel flexibility is the defining engineering principle. An engine that runs on landfill gas today can run on biogas tomorrow and on a renewable gas blend the year after. The asset does not change. The fuel pathway does. This is what makes distributed generation based on fuel-flexible engines a long-term infrastructure asset rather than a fixed-fuel commitment.
What is combined heat and power and why is it relevant to data centers?
Combined heat and power, also known as CHP, is the simultaneous generation of electricity and useful heat from a single fuel source. It is significantly more efficient than generating electricity alone because the heat that would otherwise be wasted is captured and used. In data center applications, CHP is relevant because it can provide onsite prime power generation with high availability, reduce dependence on grid supply, and in some configurations support cooling loads through combined cooling heat and power arrangements. CHP systems based on gas engines have decades of operational track record in mission-critical industrial environments.
What does resilience mean as an engineering outcome in distributed energy systems?
Resilience is not a component specification. It is a systems outcome. It requires thinking about fuel storage and supply chain vulnerability, designing for maintainability under difficult rather than optimal conditions, understanding how a system behaves when something unexpected happens rather than only when everything works as planned, and designing with the full operational life of the asset in mind rather than its commissioning performance alone. These principles were developed through operational experience in remote and constrained environments where the consequences of getting resilience wrong were immediate and serious.
Why are grid constraints driving interest in onsite distributed generation for data centers?
Grid connection timelines in many markets have extended significantly, in some cases deep into the latter part of the decade. When grid timelines extend beyond investment horizons, when power density exceeds what traditional infrastructure was designed for, and when the cost of downtime becomes large enough that reliability is a strategic priority rather than a specification, the case for onsite distributed generation changes fundamentally. It moves from being a backup consideration to being a primary infrastructure decision. This shift was visible in the data center sector from around 2016 onward and became industry-defining from 2023.
What is the Cogen World Coalition and what role does it play in the energy sector?
The Cogen World Coalition is an international organisation representing the combined heat and power and cogeneration industry. It advocates for policy frameworks that recognise the efficiency and resilience benefits of distributed CHP generation, engages with governments and regulators on energy transition strategy, and connects the global CHP community across industrial, commercial, and infrastructure sectors. The coalition became increasingly central to the data center energy conversation as CHP and distributed generation moved from background considerations to primary infrastructure options for AI-era facilities.
Why does experience in developing markets matter for understanding AI data center power challenges?
In markets where grid infrastructure is constrained, unreliable, or absent, distributed generation is not a supplement to central power. It is the infrastructure. The engineering questions that arise in those environments, about fuel supply chain vulnerability, system behaviour under unexpected conditions, maintainability under difficult operating conditions, and long-term asset performance, are the same questions now emerging in developed markets as AI data center growth collides with grid constraints. The difference is context, not substance. Operational experience from constrained markets provides a practical education in resilience that laboratory conditions and stable grid environments cannot replicate.