Many Canadians may be surprised to learn that the country already hosts more than 300 data centres, including some built specifically for artificial intelligence. Microsoft has added to this growth with a major investment that includes four large-scale data centre campuses near Quebec and Ontario, built primarily to support AI workloads.

For facility managers — particularly those overseeing large properties or multi-building campuses — these developments have real operational consequences. Data centres place heavy demands on electrical grids and water supplies, which can affect the infrastructure and resources that other facilities depend on. They also tend to trigger new policies and regulatory changes that managers will need to navigate.

Three areas deserve close attention: electrical demand, water consumption, and the evolving policy landscape. Each is covered in more detail below.

Electrical needs

Data centres are enormous consumers of electricity, often exceeding the power needs of entire surrounding communities. A typical facility draws between 50 and 100 megawatts, enough to power up to 1,000 homes a year.

Meeting that demand usually requires significant grid upgrades, which may bring with them new rules around peak-load management. This refers to situations where the grid delivers electricity at full capacity.

When this happens, it can mean power limitations at certain times of day and/or time-of-use pricing where electricity costs rise during high-demand windows, such as from 11 a.m. to 2 p.m. The goal of time-of-use pricing is to encourage reducing electrical demand at certain times of the day, but facility managers may be limited as to how much they can reduce consumption. Constrained energy availability can also delay building upgrades and construction projects.

Water

Data centres consume water at a scale that can rival or exceed neighboring communities, and facility managers should expect increased competition for municipal supply. That pressure tends to push up both water delivery and wastewater treatment costs.

As with electricity, local infrastructure often isn’t built to handle these volumes and because data centres rarely fund the necessary upgrades themselves, existing customers typically absorb the cost through higher rates.

When it comes to water, this situation is accelerating a shift toward more water-efficient fixtures. In most facilities, restrooms are the single largest source of water use, so this is a target area in which to reduce water consumption.

For instance, formerly standard 1.6 gallon-per flush toilets should no longer be installed, they should be replaced with 1.28-gallon models, which are now becoming the industry standard.

Further no-water or waterless urinals are now considered the most efficient option available when it comes to urinals, which use anywhere from one to three gallons of water per flush. Beyond the fact that no-water is necessary for these urinals to operate effectively, installation costs are often considerably less with these urinals, making them more cost effective as well.

Policy and regulatory changes

Not every Canadian municipality will introduce new rules in response to data centre growth, but many will, and facility managers should stay ahead of what’s coming.

Zoning regulations may change to confine data centres to industrial and commercial areas, with buffer zones established to limit their impact on residential neighborhoods. Noise rules and policies are also likely to emerge. Though rarely discussed, data centres can generate up to 105 decibels of noise — comparable to a jet engine — from their cooling systems, generators, and gas turbines.

Water-use restrictions are another possibility. While these are traditionally associated with drought, some municipalities may impose consumption limits by time of day regardless of local rainfall conditions.

What facility managers should do now

The impacts discussed above represent only a portion of the changes that large, resource-intensive data centers can impose on commercial properties and entire communities. Responding effectively will require deliberate planning and a proactive strategy – and that work needs to begin now.

Artificial intelligence, which depends on these data centres, is expanding globally at an extraordinary pace. What was considered “experimental” only a year ago has rapidly become foundational. By 2026, AI is predicted to become a decisive technology embedded in the structure of the global economy. In many cases, national economies and critical operations increasingly depend on it to function.

For facility managers, this shift underscores the urgency of preparing for higher energy and water demands, evolving regulations, and long-term infrastructure constraints. Those who plan early and adapt strategically will be far better positioned to manage risk, control costs, and maintain operational resilience.

Additionally, key steps to take now include the following:

• Set benchmarks for your facility’s electricity and water use. If you decide to cut consumption, these benchmarks will help track your progress. Also, consider installing multiple metering systems (electricity and water) in different areas of the facility. This allows managers to identify which areas are improving and where additional efforts are needed.
• Connect with local utility companies, government officials, and zoning committees to stay informed about upcoming data centre plans. This ensures managers can protect their facility’s interests and influence community decisions.
• What plans and strategies will be implemented in your facility if local utilities impose restrictions of time-of-use electricity pricing, as discussed earlier? Will lighting need to be dimmed or HVAC systems adjusted? These scenarios should be discussed and planned now, making them easier – and likely more cost-effective – to address.
• Consider all building mechanical systems (cooling systems, water heaters, lighting, restroom fixtures, and more) to identify where upgrades can reduce both electrical and water use and help adapt to changes driven by data centres.

What this all comes down to is planning. Former president and military leader Dwight Eisenhower is often quoted as saying, “Plans are nothing. Planning is everything.” What Eisenhower meant is that the process of planning for various scenarios helps managers think through challenges, identify risks, clarify priorities, and better understand available resources.

Such planning, sooner than later, will enable managers to more effectively address the AI and data centre challenges sure to come.

Klaus Reichardt, founder and CEO of Waterless Co. LLC in Vista, California, is a recognized authority on water conservation. Since 1991, he has led the company to innovate plumbing products like the Waterless No-Flush urinal, which operates entirely without water.