The Small Things That Bring Data Centres Down

Why the pipes nobody thinks about matter more than you'd expect

When people talk about data centre downtime, the conversation usually goes straight to the big stuff: power systems, cooling capacity, cybersecurity threats. And fair enough -these are serious risks that deserve attention.

But here's what often gets missed: many outages don't start with a server failure or a sophisticated attack. They start with something far more mundane. A pipe joint that's been quietly degrading. A drainage system that nobody thought twice about. The unglamorous infrastructure that everyone assumes will just work.

Until it doesn't.

When the background systems move to the foreground

Think about the pipe systems running through a data centre. They're carrying cooling water, condenser loops, treatment chemicals - keeping everything at the right temperature, the right pressure. They're dealing with constant thermal expansion and contraction, vibration from equipment, and the reality that data centres never stay static. They grow, they change, they get modified.

Most pipe failures don't announce themselves dramatically. They start quietly:

• A joint that handles thermal cycling just fine for the first few years, then gradually weakens
• A material that passes all the tests at installation but ages in ways nobody predicted
• A system designed for today's load that can't flex when tomorrow inevitably arrives

These aren't really technology failures. They're decision failures - choices made early on about materials, installation methods, or system design that seemed reasonable at the time but had hidden costs down the line.

And when a leak does happen? In a mission-critical facility, even a small amount of water in the wrong place can cascade quickly. It reaches electrical systems, plant rooms, control panels. What started as a minor leak becomes an incident.

Why pipework deserves better than "good enough"

There's a tendency to treat pipes as commodities - interchangeable, simple, low-risk. Pick something off the shelf that meets the basic spec and move on.

But that's not how it works in practice.

A pipe system that'll still be reliable in ten or twenty years needs:

• Materials chosen for the actual operating conditions - not just what's on paper
• Joints that maintain their integrity through decades of thermal movement and vibration
• Consistency across the system so maintenance teams aren't constantly adapting to different standards
• The flexibility to handle future modifications without becoming a liability

This is where IPS's partnership with AGRU Kunststofftechnik GmbH comes in. We supply engineered thermoplastic pipe systems that use homogeneous welded joints - essentially eliminating the mechanical weak points you get with traditional jointing methods. It's not about being clever for the sake of it. It's about creating systems that behave predictably, year after year.

Because in a data centre, predictability is reliability.

Designing for reality, not just the datasheet

One of the biggest gaps in pipe system design is the difference between what's assumed and what actually happens.

Data centres aren't static. They expand. Cooling loads increase. What was supposed to be temporary becomes permanent. The pipework installed in phase one ends up supporting demands it was never meant to handle.

When you design a system to barely meet the minimum spec at commissioning, you're essentially guaranteeing it'll become fragile over time.

That's why we work with designers and contractors to make sure pipe systems are:

• Selected for real-world operating conditions, not just theoretical limits
• Designed to accommodate thermal expansion without fighting itself
• Standardised to reduce complexity and the chance of errors during maintenance
• Built to support future changes without compromising what's already there

Good design doesn't prevent change - it absorbs it.

Getting it right from the start with AGRUCAD Digital Libraries

Here's an uncomfortable truth: most fluid system risks get locked in long before anyone picks up a wrench. Once the materials are specified, the layout is set, and the jointing methods are chosen, your options for reducing risk narrow dramatically.

Through our partnership with AGRU, we help designers access AgruCAD Digital Libraries during the planning phase. This means you can:

• Specify the right components early, when it's easy to change them
• Eliminate ambiguity that leads to substitutions and compromises later
• Coordinate better between mechanical and drainage systems
• Catch design decisions that would have become operational headaches

Better information earlier means fewer compromises later. And fewer compromises mean systems that actually work as intended.

Drainage: the unglamorous hero

Nobody gets excited about drainage. It rarely comes up in conversations about uptime.

But it absolutely should.

Data centre drainage systems have an unusual job. They're not dealing with constant, predictable flow. They need to handle:

• Regular condensate and purge water
• The occasional leak that nobody planned for
• Maintenance flushes and cleaning cycles
• Abnormal conditions that only happen when something else has already gone wrong

When drainage is undersized or poorly planned - when it's treated as an afterthought - water has nowhere safe to go. And when that happens, it will always find the most damaging path available.

We treat drainage as critical infrastructure. Because that's what it is.

Making drainage decisions easier with Kessel Smart Select

To help engineers get drainage right at the design stage, we support tools like KESSEL Drainage Technology Smart Select. It helps you:

• Size drainage and backwater protection correctly for actual conditions
• Select products based on real flow scenarios, not generic assumptions
• Identify hidden capacity issues before they become problems
• Design for abnormal events, not just normal operation

Good drainage is drainage you never think about. That's the goal.

The pattern behind preventable downtime

Most downtime connected to pipes, fittings, and drainage has something in common: it was predictable. It was preventable. And it started with small decisions that didn't feel critical at the time.

Someone chose a slightly cheaper material. A detail got simplified during value engineering. A future expansion wasn't fully considered. None of these felt like major risks in the moment.

Reliability doesn't come from adding redundancy at the end of a project when you realize things are fragile. It comes from engineering discipline at the beginning—from thinking through material selection, joint integrity, drainage performance, and design clarity before concrete gets poured and pipes get installed.

That's what IPS Flow Systems focuses on: reducing risk where it actually starts. Quietly, consistently, and well before systems go live.

Because the best infrastructure is infrastructure you never have to think about.