Selling a generator can produce revenue after contractual delivery. Selling dependable power requires funding the equipment first and recovering that capital over years. Partners Group’s planned investment in Britain’s AVK Power Solutions brings that business-model change to data-center infrastructure. It can improve revenue visibility while putting more obligations on the balance sheet.[1]
On August 6, Partners Group announced a majority acquisition on behalf of clients, with more than $1 billion of initial equity planned alongside debt and management retaining a minority interest. The funding supports growth and is not a disclosed purchase price or enterprise value. Earnings and net debt needed for an acquisition multiple are unavailable.
In this article7 sections
Installed history is not owned operating capacity
AVK designs, installs and maintains backup generation, bespoke microgrids and modular electrical systems. It has delivered more than 20,000 projects and installed approximately 3.5GW. This establishes engineering experience, not ownership of every installed asset or continuous operation of all capacity.[1]
If equipment has been sold, the customer owns its generation economics and AVK may retain maintenance work. Ownership under a service contract creates a different capital and residual-value exposure. Dividing planned equity by historical installations would combine growth funding with assets that may belong to customers.
The company also has more than 2GW of advanced behind-the-meter opportunities. The announcement does not define how many have binding customers, financing or permits. The pipeline describes business development rather than operating supply or next-year revenue.
AVK covers Frankfurt, London, Amsterdam, Paris and Dublin—the FLAP-D markets—and emerging European locations. A regional service network may help, but fuel, permitting, tax and energy conditions differ. Repeating a technical design does not guarantee repeating its return.[1]
Customers are buying earlier operation
Partners Group identifies grid congestion as an opportunity and describes connection waits of seven to fifteen years in major markets. That is the buyer’s market assessment, not a timetable for every site. Each project needs a comparison of grid expansion with on-site generation.[1]
For a campus with customers, earlier power can bring rental or compute receipts forward and reduce idle server investment. A developer may rationally pay more for a temporary power solution. The premium is bounded by additional cash contribution after energy, maintenance, financing, emissions controls and installation.
For illustration, if starting a year earlier generates $30 million of cash contribution but on-site power adds $18 million of annual cost, the first-year benefit is $12 million before one-off installation and exit costs. These are not AVK customer figures. They show how deployment speed should be priced against customer economics.
If no committed workload exists, earlier generation may simply bring fixed expense forward. The strongest projects align customer demand, site readiness and a practical power plan. Faster supply does not itself create end demand.
Recurring fees require capital before collection
Partners Group intends AVK to build and own power assets supported by long-term offtake contracts: Energy-as-a-Service. Customers can reduce direct procurement and operating responsibilities; AVK must finance equipment, construction and the service network. The announcement does not disclose the present EaaS revenue share.[1]
Recurring receipts are not automatically more profitable than equipment sales. A sale can recover procurement costs over a shorter cycle. Service assets require investment before years of receipts. Stable operating earnings and declining free cash flow during expansion can coexist, so contract growth needs to be assessed beside funding consumption.
Contracts can allocate costs through fixed availability fees and variable energy charges. Fuel escalation, downtime deductions and minimum payments under reduced customer usage determine quality. These are analytical contract options, not claims that AVK uses a uniform template.
Usage-based payment leaves more utilization risk with the operator. Capacity-based payment shifts idle-demand exposure toward the customer but requires the supplier to maintain availability. Identical forecast revenue can conceal very different downside cases.
Long duration also does not remove credit exposure. A campus relying on one model company needs an assessment of the ultimate workload and payer. Equipment security helps only to the extent assets can be relocated, relicensed and sold at acceptable cost.
The cash requirement of a hypothetical 100MW project
Assume, solely for illustration, 100MW costs $100 million initially and requires $8 million in annual fixed maintenance and operations. An assumed annual capital cash requirement of 10% adds $10 million, before separately addressing tax, financing and end-of-life capital recovery. Fully contracted fixed charges would need to cover at least $18 million a year, or $180,000 per MW, with fuel additional.
| Independent assumptions | 100MW contracted | 70MW bearing fixed costs |
|---|---|---|
| Annual fixed and capital-cash requirement | $18m | $18m |
| Required payment per contracted MW-year | $180,000 | Approximately $257,000 |
| Unit fixed-charge change | Base | Approximately +43% |
| Excluded | Fuel and other variable costs | Fuel and other variable costs |
If only 70MW is contracted while asset cost and fixed expense remain unchanged, the same $18 million becomes approximately $257,000 per contracted MW, around 43% higher. This measures contracted-capacity sensitivity, not AVK pricing or realized profit.
At 80% average operation, 100MW generates approximately 700,800MWh a year. A $10/MWh increase in fuel and other variable costs adds roughly $7.01 million annually. Without pass-through, that consumes about 70% of the illustrative $10 million capital cash requirement.
The calculations isolate two different risks: unsold capacity increases the burden of fixed cost, while variable-cost inflation reduces energy contribution. Minimum offtake and fuel pass-through can make a project more infrastructure-like. A long-term contract alone does not establish stability.
Debt reduces initial equity per project but creates scheduled interest and principal. Construction before customer commitment or delayed usage does not automatically suspend those payments. Without disclosed borrowing terms, an asset-level example cannot produce a defensible equity IRR.
Operating performance determines replication
Frequent or continuous operation changes maintenance, spare-parts requirements and the consequences of failure. Redundancy, batteries, control systems and field staff have to keep service available when a generator fails. A growing installed base requires a service network capable of supporting it.
Compatibility with alternative fuels does not establish affordable local supply. Gas connections, fuel quality, emissions controls and permitted operating hours still constrain projects. A generator’s advertised starting speed also does not establish full-campus recovery performance.
On-site power can reduce immediate dependence on grid expansion without reducing long-term grid needs. Customers may retain backup connections or change the generator’s role once utility supply arrives. The operator needs a commercial plan for that transition, after the highest-value period of early supply ends.
Scale benefits may come from purchasing, service scheduling and financing experience. Nearby sites can share parts and engineering teams. Expansion into distant markets with different permits can instead raise the cost of delivering every contracted megawatt.
Customer credit and exit value determine a long contract’s worth
Power-as-a-Service replaces an initial equipment purchase with payments over time, but duration alone does not guarantee recovery. The counterparty might be an established cloud company or a project vehicle dependent on future financing. Both operate in data centers, yet their payment capacity differs. The actual obligor, parent support, advances and repossession rights matter.
Movable generation may preserve more exit value than entirely dedicated infrastructure. Removal, transport, new permissions and idle time still cost money. Pipelines, civil works and custom connections can be much harder to recover. A model should not assign movable-engine residual value to every dollar invested.
Temporary supply while waiting for a grid connection also differs from permanent primary generation. Early grid availability raises the question of termination compensation; a delayed connection raises the price of an extended service period. Those provisions determine capital recovery and how the benefit of earlier operation is divided.
Additional capital helps AVK accept larger projects, but growth can also lock more cash in equipment and receivables. A repeatable model should preserve adequate advances, controlled construction and collection as volume grows, rather than require progressively greater leverage with every long-term contract.
What would make the investment work
The positive case combines durable customer demand, AVK’s installation and maintenance experience, and capital that allows the company to retain assets and customer relationships. Well-matched duration, minimum payments and fuel adjustments can allow mature projects to finance subsequent growth.
The counter-case does not require AI demand to disappear. Faster grid connections, postponed campuses, more expensive fuel or emissions compliance, and expansion beyond service capacity can all compress returns. Power scarcity becomes supplier profit only if the supplier delivers reliably at an acceptable cost.
More than $1 billion of planned equity enables expansion but also increases the capital that investors need to recover. Closing the acquisition, signing binding offtake, completing projects within budget and generating cash after maintenance and debt service are separate tests. Missing acquisition and operating financials prevent a precise valuation.
Over twelve to thirty-six months, compare contracted and operating capacity within the 2GW pipeline, EaaS mix, additional capital per project and availability. Growing operating cash without repeated equity requirements would support the transformation. Recurring revenue accompanied by continually rising capital needs would not, by itself, demonstrate shareholder value.
Sources
This website is for information and research only and does not constitute investment advice.