Sustainable data centre patent strategy: claims, supply and location
The growth of artificial intelligence (AI) and cloud computing is increasing the importance of efficient data centre infrastructure. The notorious demands for water and electricity to operate these facilities presents opportunities for technical innovation. However, obtaining commercially useful patent protection requires more than identifying a technical solution to an environmental problem. It is also important to consider how an invention will be supplied, where it will be used, and whether infringement could be identified.
Where is innovation occurring?
Cooling is one important area. Evaporative cooling consumes water as part of the heat removal process, whereas closed-loop arrangements recirculate coolant. Developments intended to reduce water consumption include improved cooling loops and systems using fluids other than water. The way heat is ultimately rejected also matters, with some developments targeting efficient heat exchangers to extract the heat from the closed loop. From a patent perspective, room for inventions might lie in the arrangement of fluid pathways, the interaction of cooling components, or the way circulation is controlled, rather than in an entirely new component.
Electricity consumption presents a different set of challenges. A large facility supporting cloud-based AI workloads may have a substantial, relatively sustained power requirement, while a smaller edge data centre serving nearby users may experience greater fluctuations with demand. Improvements may therefore concern both the computer hardware used to process a workload and the techniques used to manage operation. For example, specialised processing hardware may reduce the energy needed for a given task, while a scheduling technique could defer non-critical workloads to quieter periods to smooth peaks in power demand.
Protecting the right part of the infrastructure
A data centre is not generally supplied as a single finished product. Equipment may be provided by several suppliers and connected on site by an operator or specialist installation contractor. A cooling invention might use otherwise conventional pumps, pipes and heat exchangers arranged in a new way. If a patent claim requires the entire installed arrangement, the claimed configuration may only come into existence at the data centre. Supplying an individual component would not necessarily directly infringe that claim. This can be disadvantageous where the inventive contribution lies in one or more sub-components, because those components may be manufactured, bought and sold without falling within the scope of a claim directed only to the complete installed system.
Enforcement may then depend on proving that the components have been combined and operated in the claimed arrangement, rather than addressing their manufacture or supply at an earlier stage. This can be particularly difficult once the data centre has been constructed, since the relevant equipment may be located in a secure or inaccessible facility, making it difficult to inspect the installation or determine whether the patented combination is present. These practical difficulties can reduce the commercial value of protection drafted solely around the completed data centre.
It is therefore worth considering whether the inventive concept can be protected at the level of individual subsystems, controllers or other separately supplied apparatus, which can be manufactured and protected independently. For a cooling-loop invention, for example, it may be possible to define the relevant fluid connections and control features without requiring the servers themselves to form part of the claimed apparatus. This can avoid making protection unnecessarily dependent on the particular computing equipment being assembled with the cooling system.
Applicants should bear in mind that this is not simply a matter of removing features to make a claim broader – the claim must still define a new and inventive technical contribution. Where a particular hardware component provides that contribution, apparatus claims may offer a useful opportunity to address manufacture and supply before installation. Where the contribution lies only in the assembled arrangement, system or method claims may remain essential. We recommend discussing the intended supply chain with a patent attorney as early as possible.
Where should protection be sought?
Patents are territorial, so the physical location of a data centre can be particularly important for an invention implemented only after equipment has been installed. Protection in a supplier’s home country may provide limited assistance if the claimed system is assembled and operated elsewhere. Conversely, a patent covering a separately supplied inventive apparatus may be relevant in a country where that apparatus is made, even if the eventual data centre is elsewhere. The relevant countries can therefore differ depending on what is claimed.
Country selection should reflect likely deployment locations as well as manufacturing and supply markets. Larger economies such as the United States may be an important candidate for many businesses, but it should not be the only location considered. There are some noticeable trends to deploy data centres to more remote high-latitude locations, such as the Nordics, due to more abundant renewable power and natural cooling possibilities. Location preferences and infrastructure investment can change, so filing decisions should be informed by the applicant’s customers, competitors and expected markets rather than assumptions about where data centres, or the companies that operate them, are traditionally found.
In Europe, there are several ways of obtaining effective patent protection. Upon grant of a European patent, the applicant has a choice of validating the patent in whichever countries are most useful for a particular business strategy. Alternatively, with the introduction of unitary patents, applicants can obtain more cost-effective protection across all participating countries. This can be particularly useful, for example, where data centres could be situated in such a variety of locations. The appropriate combination of unitary protection and national rights will depend on the geographical coverage required and the applicant’s wider strategy.
Cross-border operation raises a further issue where, for example, a customer’s device is in one country and the data centre providing the service is in another. We discussed these considerations in our article Cross-border computer systems: patent infringement across multiple jurisdictions.
Practical advice
For inventions with an environmental benefit, the UK Intellectual Property Office’s (UKIPO’s) “Green Channel” offers a route for accelerated processing, which may help obtain an early patentability assessment during the priority year before major overseas filing decisions are made. More generally, a useful starting point is to identify the technical contribution, the equipment or process that embodies it, who supplies or operates that equipment, and where those activities will occur.
Sustainable data centre technology can create valuable opportunities for patent protection, but there is no single filing or claim strategy suitable for every development. We recommend seeking advice from a patent attorney familiar with the technology and explaining how the invention is intended to be commercialised to reach an appropriate claim scope and international filing strategy.
Related article
Cross-border computer systems: patent infringement across multiple jurisdictions: https://dycip.com/computer-systems-cross-border

