CASE STUDY
In the race to power AI, data centres are hitting new heat densities and sharper water constraints. Cooling can no longer be a box on a CapEx sheet – it has to be a living programme that proves uptime, efficiency and sustainability in real time. Few people sit closer to that shift than Dr Nevin Sant, Vice President, RD & E – Global Light Water Sector, High Tech and Power Industries at Ecolab. For over 11 years, she has led RD & E across high-stakes environments from data centres and microelectronics to pharma and heavy industry, along with corporate roles taking ideas from lab to pilot to global rollout.
In this conversation, Dr Sant explains why Cooling-asa-Service is replacing standalone kit with system-level orchestration, with chemistry, mechanics and software working as one. She explains how clear SLAs and realtime controls translate into better WUE and PUE, fewer unplanned outages and cleaner ESG evidence.
In addition, she unpacks Ecolab’ s co-innovation model with operators, the push for standardisation through the Open Compute Project – especially around direct-to-chip liquid cooling – and the role of digital backbones like 3D TRASAR in turning telemetry into action. We also explore how to keep service quality consistent across regions while tuning for local realities such as water scarcity. The throughline is simple: treat cooling as an outcome engine, not an asset, and design for measurable performance from day one.
Can you start by introducing yourself and your role at Ecolab?
“ Throughout my 11 years, I’ ve moved from hands-on R & D into system-level programmes where chemistry, mechanics and software meet. That evolution mirrors where the sector is heading – away from treating cooling as isolated kit and towards Cooling-as-a- Service that orchestrates the whole chain in real time.
“ The aim is rigorous and simple: use the least water and power while holding tight thermal envelopes, which shows up in better WUE and PUE and a lower cost to compute. The same instrumentation that tunes efficiency also protects uptime – early signals in coolant health or heat-exchange drift are surfaced before they become incidents – and because the programme is data-first, customers get credible evidence against sustainability targets with the audit trails boards expect. All of these gains reinforce one another.”
How is the industry shifting towards Cooling-as-a-Service( CaaS) and how does this impact how customers view cooling infrastructure today?
“ The shift is from CapEx-thinking to outcome-thinking. Instead of buying equipment and measuring against a datasheet spec, operators contract for performance – uptime, thermal stability, asset life and resource efficiency across water, energy and labour – under clear SLAs.
“ CaaS lets us orchestrate equipment, chemistry and controls as one living system. That orchestration is what enables real-time optimisation: we apply just enough water and power to meet thermal targets with headroom, which moves WUE and PUE the right way and trims cost-to-compute. Because monitoring is continuous, we catch small deviations early and avoid the sort of thermal-related shutdown that carries revenue and reputation risk. And because the whole programme is instrumented end-to-end, sustainability reporting becomes a by-product of good operations rather than an afterthought.”
Can you explain the rising role of customer collaboration and coinnovation in building next-generation service-based cooling models?
“ Innovation cycles have compressed to months instead of years, and the only way to move at that speed is to co-design with customers around the outcomes they care about. A good example is our direct-to-chip coolant-health monitoring. We built it shoulder-toshoulder with operators, ran live pilots, tightened the sensing and service playbooks, which meant we transitioned from idea to launch in under eight months.
“ The value is not the gadget – it’ s the system: when coolant properties, flow and thermal behaviour are monitored and acted on in real time, you hold temperature more precisely while using fewer resources, so WUE and PUE improve and uptime hardens.
“ In addition to co-inventing with customers through clear partnerships, we spend significant time with industry bodies such as the Open Compute Project to drive standardisation and document best practices for data-centre cooling – particularly direct-to-chip.
“ That focus on common methods and metrics is what brings consistency across vendors and sites, reduces risk in closed-loop liquid systems and helps the whole ecosystem orchestrate cooling as a programme rather than a pile of parts. It also means sustainability
www. intelligentcio. com INTELLIGENTCIO NORTH AMERICA 15