And here's what solving it actually looks like.
The compensation loop
Most data centers reach a point where operations teams are doing something that shouldn’t be their job: compensating for a design that no longer fits what the facility is being asked to do.
It shows up in different ways. Cooling running harder than it should. Power headroom tighter than anyone’s comfortable with. Hotspots appearing in places that don’t make obvious sense. Workarounds that started as temporary measures and quietly became permanent.
None of these are operations failures. They’re design misalignment, and the distinction matters, because the solution isn’t operational. You can’t maintain your way out of a design problem.
The facilities managing this best aren’t the ones with the biggest ops teams or the most sophisticated monitoring. They’re the ones that have gone back to the design layer and addressed the root cause.
This is the conversation we have with customers every day. And it’s the one worth having before the compensation starts costing you more than a targeted upgrade would.
Why design is no longer a one-time phase
Data center design used to be a discrete event. You designed the facility, you built it, and then you operated it. Design was finished when construction started.
That model doesn’t hold any more, and the reason is straightforward: the workloads facilities are being asked to support have changed faster than the facilities themselves.
A facility designed for 5kW average rack density a decade ago is now being asked to support 20kW, 40kW, or in some AI configurations, significantly more. The electrical and mechanical systems, the airflow architecture, the power distribution, none of these were sized for that load. They were right for their time. They’re not right for now.
Design, in this environment, has to become an ongoing requirement rather than a historical event. Not continuous redesign, but periodic, structured reassessment of whether the facility’s design still matches its operational reality. And when it doesn’t, a clear pathway to address the gap.
What good design engagement actually looks like
When customers come to us with infrastructure challenges, the starting point is always the same: understand the gap between design intent and current operational reality before recommending anything.
That sounds obvious. In practice, it’s rarer than it should be. A lot of infrastructure decisions get made on the basis of operational observation, something isn’t working as expected, rather than design analysis. The result is solutions that address the symptom rather than the cause.
Our process starts with a structured assessment: current load vs. design capacity, thermal performance against modelled parameters, power distribution architecture against actual usage patterns. This gives us a clear picture of where the misalignment is, and how significant it is.
From there, the options become much clearer. In most cases, a targeted upgrade, focused on the specific systems that are out of alignment, is significantly more effective, and significantly less disruptive, than broader interventions.
The core capabilities we bring to this work:
- Mechanical and electrical engineering, in-house, not subcontracted
- CFD modelling across white space, grey space, and external plant
- BMS and Energy Management System integration
- Power and cooling upgrade design and delivery
- Liquid-to-chip cooling for high-density and AI workloads
A note on CFD — why it matters more than it used to
Computational fluid dynamics modelling has become a standard part of how we approach any thermal challenge. The reason is simple: the behaviour of air, or liquid, in a high-density data center environment isn’t intuitive, and getting it wrong is expensive.
CFD lets us model what’s actually happening in a space before we make physical changes to it. In an upgrade scenario, that means validating that the proposed solution will perform as intended before anything is installed. In a design scenario, it means eliminating uncertainty about thermal performance before equipment is ordered.
We apply CFD across three domains: white space (active compute environment and airflow patterns), grey space (mechanical plant and cooling infrastructure), and external facility (site-level airflow and heat rejection). The output isn’t a visualisation, it’s design validation. Confidence that the solution will work.
For customers deploying AI workloads at high density, this isn’t optional. The thermal behaviour of a 40kW rack in a space designed for 8kW is not something you want to discover after deployment.
Design through to operations, why the handoff matters
One of the things that consistently creates problems in data center projects is the handoff between design and operations. When these are handled by different organizations, or by the same organization with poor continuity, decisions made at the design stage don’t land correctly in operations.
Designers who have never operated a data center make decisions that are correct on paper but create operational headaches in practice. Equipment that looks right in a specification doesn’t perform as expected in a live environment because the operational context wasn’t understood at the design stage.
Our approach spans design, procurement, and operational planning, and critically, the design team stays engaged through to handover. The people who understand why decisions were made are available to the people who have to live with those decisions. That continuity changes the outcome. It also changes procurement. When the design team is integrated with the procurement process, long-lead items are identified early and acquisition begins while design is still being refined, rather than after it’s finalised. In a market where lead times for critical equipment can stretch beyond twelve months, that’s not a nice-to-have. It’s how you hit your timeline.
Where to start
The easiest entry point into a design engagement isn’t a full project. It’s a consultation, a structured conversation with our design team about your current position, your operational challenges, and your forward requirements.
For many customers, that conversation is valuable in itself: a clear view of where the design gaps are, what the options look like, and what a realistic remediation pathway involves. It shapes capital planning conversations and gives you something concrete to take to internal stakeholders.
For others, it’s the start of a project. Either way, the right time to have it is before the operational pressure is acute — when the options are still broad and the timeline still has room.
Our current entry points:
- Operational roadmapping — structured assessment of current state vs. future requirements
- Reliability improvement — targeted identification and remediation of design-driven risk
- Design planning and validation — CFD modelling and engineering review of proposed changes
- Full upgrade delivery — mechanical, electrical, BMS, power and cooling
Request a design consultation with our team. We’ll come prepared with questions, not a sales deck.