Datacenter Management Practices

Everyone's chasing data center land. Almost everyone is missing the real constraint. It's not fiber. It's not even land. It's power. U.S. Interior Secretary Doug Burgum said at the Prologis conference: "To win the AI arms race against China, we've got to figure out how to build these artificial intelligence factories close to where the power is produced, and just skip the years of trying to get permitting for pipelines and transmission lines." Translation: The next generation of data centers won't be built where the land is cheap. They'll be built where the power is available. Three implications for dirt investors: 1. Nuclear Proximity = New Premium: Amazon already signed deals with Dominion Energy near the North Anna nuclear power station in Virginia and expanded partnerships with Talen Energy at the Susquehanna nuclear plant. Sites within transmission distance of existing nuclear facilities just became exponentially more valuable. 2. Warehouse Conversions Accelerate: If Prologis is eyeing their 6,000 buildings for data center conversion, every industrial site with surplus power capacity needs re-evaluation. What looks like a struggling warehouse today might be a data center tomorrow. 3. Grid Capacity > Geographic Desirability: Constellation Energy CEO Joseph Dominguez noted that data economy customers "want to run their systems 24-7" with "firm pricing so that they know the price for energy for 20 years". Long-term power contracts are becoming the new land entitlements. But here's what nobody's talking about: The same power constraints driving this opportunity are also creating massive project risks. According to a recent CoStar analysis, data centers will account for up to 60% of total power load growth through 2030. But there's a timing mismatch: data centers take 2-3 years to build, while power system upgrades take 8 years. That gap is forcing developers to either wait or find sites with existing capacity. The Community Resistance Factor Data Center Watch estimates $64 billion in data center projects were blocked or delayed over a recent two-year period. There are now 142 activist groups across 24 states organizing against data center development. Northern Virginia alone-the nation's largest data center market-has 42 activist groups fighting projects. Reasons cited: water consumption, higher utility bills, noise, decreased property values, loss of open space. Translation for land investors: Sites with existing power capacity + community support just became exponentially more valuable than sites with just land and zoning. The power infrastructure thesis isn't just about finding available capacity. It's about finding that capacity in counties that actually want data centers. Not every market will roll out the welcome mat. Are you evaluating community sentiment alongside power infrastructure access?

Global data centre power demand is expected to reach 130 GW by 2028. Growing at 16% per year. But infrastructure isn’t built in CAGR charts. It’s built with copper, transformers, and time. You can’t scale compute if your logistics don’t scale first. Transformer lead times now exceed 18–30 months High-purity copper fluctuates 20–40% annually PDUs and switchgear bottleneck at tier-2 fabs in Malaysia or Taiwan. In the GCC, grid access and hardware supply are the real blockers. You can’t fix this reactively. Therefore, do this ➝ Engage EWEC and ADPower at design freeze (+12 months) ➝ Tie permits to confirmed upstream generation capacity ➝ Source servers from Dell Technologies, Inspur Group, or others ➝ Use dual-rated transformers with voltage-class fallback mapped ➝ Lock PDU/UPS supply with Vertiv or Schneider Electric ➝ Buffer SKUs at Jebel Ali, King Abdullah Port, and Chennai ➝ Hold 90+ days of bonded spares under warehouse agreements ➝ Use Foxconn India, Tata Electronics, or Modon for sub-assembly ➝ Pre-contract backup generation with Cummins Inc. or others. ➝ Track LME copper pricing and feed into your BOM risk models ➝ Enforce fallback mapping in ServiceNow across critical components Model power constraints ⤦ Simulate dispatch curves for NEOM's wind and solar resources. ⤦ Map load by rack, season, and cooling profile. ⤦ Validate UPS and chiller curves against real site-level energy windows. Hardware gets built in Asia. Time gets lost in transit. Power gets delayed at permits. Your design isn’t complete until every component has a fallback... ...and every kilowatt has a Plan B. Save this if you’re planning anything Hyperscale.

Most people think data center site selection is about proximity to fiber and population centers. That was true 5 years ago. It's not true anymore. Here's what actually matters now: Power availability. Full stop. We've walked away from sites with perfect fiber, perfect location, perfect everything. Because the utility couldn't deliver power in a reasonable timeline. And we've pursued sites in the middle of nowhere. Because the utility had capacity and could move fast. The math has completely flipped. Proximity to end users matters less when you can build fiber. Proximity to talent matters less when you can operate remotely. Proximity to power generation matters more than anything else. Here's what we look for now: Utilities with excess generation capacity or clear path to new generation (Hint: Sometimes you have to create your own path). Regions with natural gas pipeline infrastructure already in place. Sites near existing substations with available capacity. Regulatory environments that move fast on interconnection approvals. Everything else is secondary. The crazy thing is: This is creating opportunities in places nobody's looking. While everyone's fighting over Northern Virginia and Silicon Valley, there are regions with abundant power that nobody's paying attention to. The data center map is about to get redrawn. And it's going to be drawn by power availability, not proximity to users. *Here's a picture of my favorite beach for those in colder climates 😊 *

The race for data center development is one our customers can’t afford to lose. Avoiding suboptimal locations not only saves excessive study costs but also saves time, keeping developers competitive. In a Cleveland area case study, we highlight a siting workflow that successfully qualifies and ranks over 600,000 parcels in minutes, quickly identifying fewer than 20 optimal sites, ultimately pitching 2 sites ideal for hyperscale sized development. While certain criteria are nonnegotiable when it comes to siting like sufficient withdrawal capacity, available transmission, adequate buildable acres and access to fiber-optic lines, evaluating sites on additional criteria is what really separates the exceptional sites from the rest. Enverus, Enverus Intelligence® Research

Every billion-dollar data center begins long before construction starts. But the real timeline opens the moment a site is identified… Because that’s when zoning, environmental reviews, interconnection studies, and community approvals begin. These steps decide whether a project advances or stalls. A few factors define who succeeds: 1. Permitting clarity: Jurisdictions with predictable zoning and environmental timelines. 2. Interconnection feasibility: Utilities able to commit capacity without multi-year studies or upgrades. 3. Equipment procurement: Transformers, switchgear, generators, and cooling with long lead times locked in early. 4. Commissioning discipline:  System-level tests that validate redundancy, load transitions, and operational readiness. Developers that deliver on schedule have a structural advantage. Tenants plan capacity to the month. Lenders price certainty. Missed milestones cost far more than construction overruns. The model is shifting toward standardized designs, modular electrical rooms, early procurement, and deeper partnerships with utilities and regulators. Data centers are no longer built around steel and concrete they’re built around timelines, permits, and equipment availability. Whoever controls those variables controls delivery. Read the article below #datacenters

🇮🇳 India’s Data Centre Growth – Snapshot Current capacity: ~1.2 GW Projected by 2030: ~5 GW Required Investment: ~USD 22 billion (Source: Colliers India) Current Hubs: Mumbai, Chennai, Delhi-NCR Emerging Hubs: Hyderabad, Coimbatore, Pune, Ahmedabad --- 🚀 Drivers of Growth AI and Generative AI Workloads Cloud and Edge Computing 5G rollout and IoT expansion Digital India push (e-governance, UPI, ONDC, etc.) Data localization mandates (DPDP Act 2023) Hyperscale demand from global tech giants --- 🧠 Strategic Advice for Stakeholders 1. Investors & Developers Diversify Geography: Mumbai is saturated; invest early in Hyderabad, Coimbatore, Pune, and Kolkata where land and power are still affordable. Colocation vs Hyperscale: Develop flexible colocation models catering to Tier-II startups and enterprises alongside hyperscale modules for cloud majors (AWS, Azure, Google). Green Data Centres: Focus on renewables, waste heat recovery, and AI-powered cooling to reduce OPEX and meet ESG commitments. Land Banking Now: Acquire land near RE power corridors or upcoming substations to mitigate future access and regulatory delays. 2. Power Infrastructure & Utility Players Build Dedicated Power Corridors: Ensure redundant and resilient grids with Tier IV reliability. Explore Captive RE Models: Enable direct RE connectivity (solar/wind farms in Gujarat, Rajasthan, Karnataka). Battery Storage Systems: Plan early for BESS (Battery Energy Storage Systems) to manage grid stability for AI workloads. 3. Government & Urban Planners Single Window Clearances: Fast-track environmental, zoning, and connectivity clearances. PPP Models for Tier-II Cities: Offer plug-and-play data centre parks with built-in utilities and dark fibre. Skill Development: Launch skilling hubs in electrical, mechanical, BMS, IT, and HVAC specific to data centres. 4. Telecom & Connectivity Providers Expand Redundant Fibre Rings: Enable low-latency links for AI and real-time analytics demands. Edge Data Centre Networks: Invest in micro data centres closer to users, especially in Tier-II/III towns. --- 🏗️ Big Players Making Moves AdaniConneX (Adani + EdgeConneX JV): Planning 1 GW across India, including Chennai, Noida, Hyderabad. Reliance Jio: Investing heavily in cloud + AI infrastructure, with new green data centre parks expected. NTT, STT GDC India, CtrlS, Web Werks, Sify: All expanding footprint or forming RE-linked data centre clusters. --- ⚡ Key Challenge: Power Availability & Reliability AI workloads may require up to 5–10x more power per rack Grid capacity expansion is not keeping pace in some regions. Developers should consider on-site substations, gas-based backup generation, or green open access models.

Where Would You Build a Data Center? 🚀 If you could put a data center anywhere, where would it be? Some companies have taken that question very literally and explored wild locations: 🌊 Underwater – Microsoft ‘s Project Natick tested submerged data centers for natural cooling and sustainability. 🚢 Floating on the ocean – Nautilus Data Technologies built barge-based DCs to cut real estate and energy costs. ⛰️ Inside a cave – Some data centers operate in underground bunkers for security and temperature stability. 🌖 On the Moon – Lonestar Dataholdings was to launch an Intuitive Machines mission this past week onboard the Falcon9 rocket from SpaceX carrying a “mini data center” aiming to safeguard valuable data ❄️ Antarctica – A forum discussion gave me a good laugh: “Why not just plop data centers in Antarctica for free cooling?” The answer? Power and connectivity still matter—unless you want a frozen, useless server farm. 📍 But Where You Build a Data Center is All About Balance: ⚡ Power – Cheap, stable electricity is 🔑. Hyperscalers love locations near hydro, nuclear, or solar grids. 📡 Connectivity – Close to internet exchange points (IXPs) and submarine cables for low latency. 🌪 Disaster Risks – No floods, earthquakes, hurricanes, or unstable ground (bad soil = bad idea). 🏛 Regulations & Incentives – Some governments offer 💰 tax breaks, others pile on restrictions and fees. 🛠 Talent & Logistics – Data centers do not run on their own! Remote sites struggle with hiring, and delays = 💸 lost. Sites near airports & cities win. 🌱 Sustainability – DCs consume massive ⚡ & 💧. Operators are adopting eco-friendly practices, such as using renewable energy sources and recycling waste heat, to reduce their environmental footprint. #datacenter #siteselection #AIinfrastructure #compute

With data center demand expected to double by 2030 and excess office inventory across the U.S., building conversions are becoming an increasingly compelling solution. In California, however, feasibility is multi-layered. Beyond power costs and availability, cooling, and redundancy, projects must also navigate rigorous environmental analysis through a project’s CEQA compliance. From our perspective at EPD, successful data center conversions begin with understanding environmental constraints early, not treating them as a downstream permitting hurdle. Under CEQA, adaptive reuse projects, especially those involving significant electrical upgrades or increased operational intensity may trigger detailed environmental review. Key considerations typically include:  • Energy and Power Infrastructure  Data centers are energy-intensive by nature. CEQA analysis evaluates increased electricity demand, upstream generation impacts, substation upgrades, and consistency with California’s clean energy and GHG reduction goals. Coordination with utilities and regulators is essential, particularly in constrained load pockets.  • Water Use  Cooling systems and fire suppression can significantly increase water demand. Environmental review looks at water sourcing, drought resilience, cumulative impacts, and consistency with local and regional water management plans.  • Air Quality & Greenhouse Gases  Backup generators, construction activity, and increased energy consumption raise air quality and GHG concerns.   • Noise & Vibration  Mechanical systems, generators, chillers, and 24/7 operations can impact nearby receptors. CEQA requires evaluation of operational and construction noise  • Traffic & Transportation  While data centers are not people-dense uses, construction traffic, equipment deliveries, and operational staffing still require analysis.  • Biological & Cultural Resources  In suburban, rural, or edge-of-urban locations, biological resources, wetlands, and sensitive habitats may be present. Even infill projects can raise issues related to tree removal, migratory birds, or historical resources. Data center conversions in California are absolutely viable with early-stage environmental and technical diligence to address the above multi-layered issues. At EPD, we see the best outcomes when environmental, engineering, and development strategies move forward together, especially for mission-critical infrastructure. Let's talk about your next project - https://lnkd.in/ghdf9Q_6

The AI boom is a "build-out moment." For investors and data center developers, the choices made this decade will determine if AI accelerates climate progress or becomes an unsustainable burden. With data center energy use projected to double or quadruple by 2030, sustainability is now a core risk-mitigation strategy. To future-proof your portfolio, here is what must be done before finalizing a location or design : Strategic Siting (Smart Siting) : - Analyze Hydrological Stress : Avoid water-scarce regions and favor areas with low water-stress profiles to minimize conflict with local communities. - Assess Grid Carbon Intensity : Prioritize locations with a clean electricity mix (nuclear, hydro, or windbelt states) and verify the availability of "additional" renewable capacity. - Evaluate Infrastructure Stability : Ensure local grids can handle high-density loads and support energy storage integration to balance intermittent renewables. Circular & Efficient Design : - Incorporate Advanced Cooling : Move beyond air cooling; liquid immersion or direct-to-chip systems can slash energy and water use by up to 50%. - Design for Circularity : Utilize "green" low-carbon concrete and steel, and adopt a "Design for Disassembly and Reuse" (DfDR) philosophy. - Mandate Heat Recovery : Plan to capture waste heat for district heating or industrial processes, transforming the facility into a community asset. - Perform a Cradle-to-Grave LCA : Conduct a full Life Cycle Assessment to account for embodied carbon, avoiding "environmental burden-shifting." How are you balancing the surge in AI demand with your ESG commitments? Let’s discuss in the comments. #SustainableComputing #DataCenterInvesting #GreenTech #AIInfrastructure #CircularEconomy

$2.3T data center pipeline. 68% still in pre-planning (GlobalData Q4 2025) Meaning: most of the market is still in spec-writing mode, not execution. Three gates are stalling projects: Power — With grid caps and connection timelines, power is the bottleneck. At AI training densities (often 50–120 kW/rack), parasitic load (fans, pumps) directly reduces the IT power you can monetize. Water — In many regions, WUE is becoming a permitting gate, not a KPI. Water dependency = regulator negotiation (and often schedule risk). Permitting — Noise, land use, emissions, local impact. Projects slip before procurement. Hyperscalers increasingly co-design specs early and lock long-lead supply upstream to clear these gates. So here’s the shift: architecture is the differentiator. Many specs still treat redundancy like arithmetic: N+1 as “one full extra dry cooler” is often a blunt instrument — footprint-heavy and CAPEX-intensive. But AI failures are frequently local: fan, sensor, valve, fouling. Monolithic designs turn local issues into system events. Modular changes how you engineer redundancy. You don’t improvise resilience — you design it upfront: N+1 = one extra module, not one extra unit. What that enables (when isolation and controls are designed to avoid common-mode events): Fast serviceability: isolate a module, keep the rest online. Module-level interventions can be planned around weather forecasts and load profiles, with no heavy lifts when access and spares are prepared. Granular load following: finer staging to match IT load → better part-load efficiency, tighter temperature control, lower parasitic energy. Failure containment: smaller blast radius → engineered degradation, not downtime. Second-order effect: footprint. An extra full dry cooler burns roof area and structural CAPEX. Granular redundancy protects uptime without sacrificing m² → higher heat-rejection density. The real question: which architecture clears the power/water/permitting gates and stays upgradeable as densities keep moving? At ThermoKey, we build Modular Dry Coolers and manufacture air-to-liquid aluminum microchannel heat-exchanger cores in-house, enabling geometry optimization for DC constraints: approach temperature, fan power, pressure drop, footprint, and scalability. (Patent pending) Defining a thermal concept for high-density liquid cooling? Let’s talk. #DataCenter #AICooling #Modular #LiquidCooling #ThermalManagement #WUE #PUE