Data centers are descending on Arizona, and the debate about them usually focuses on electricity (and electrical bills), water, and sometimes economic development. And sometimes health.
The Pima County Health Department recently got ahead of the data center siting curve by working with researchers from the Johns Hopkins School of Public Health to develop a Data Center Public Health Review that gives their Board of supervisors and city officials a way to look at data-center proposals through a public-health lens.
The team reviewed the available research and looked at how the building, construction, operation and decommissioning of data centers impacts community health (or could impact).
What Did They Look At?
The framework includes 21 different pathways where a data center could affect a community. Those include:
- air pollution and construction dust;
- diesel backup generators;
- pollution from power plants supplying the electricity;
- water use and wastewater;
- noise and traffic;
- waste heat and the urban heat-island effect;
- nighttime light;
- land use and loss of green space;
- grid reliability;
- electronic waste, batteries & other materials when a facility closes; and
- economic burden and social vulnerability.
PCDH then looked at Pima County’s existing health conditions like asthma, cardiovascular disease, heat illness, Valley Fever, poverty, food insecurity, urban heat and other measures of vulnerability.
There Are Still Some Unknowns
The report is not a formal Health Impact Assessment and doesn’t claim that living near a data center causes particular diseases. In fact, one of its biggest findings is how little direct research exists on the health of people who actually live near data centers.
Most of the available evidence measures or models things like air pollution, noise, heat and water use rather than directly measuring health outcomes in nearby residents.
There are also important questions that can’t be answered until the details of a specific project are known. Electricity rates are a good example.
Pima County appropriately includes household financial energy burden in their assessment. Low-income households in Pima County already spend about 15% of their income on energy costs. Higher utility bills make families choose between electricity, food, medicine and rent.
But whether a data center actually makes that problem worse depends partly on who pays for the new electrical infrastructure needed to serve it.
If a utility needs new generating capacity because a giant data center arrives, the data center should pay the full freight for the generating capacity to service their centers.
Residential customers (especially low-income families) shouldn’t subsidize or pay for the capital costs for the power plants, transmission or anything that’s needed to serve a corporate data center.
If they do – the whole equation changes as the data center then has a large impact on the social determinants of health for the entire community as people pay more for electricity to subsidize the data center infrastructure.
For utilities like APS and Tucson Electric Power, which means the Arizona Corporation Commission is a critical part of the public-health equation. The Commission approves utility rates and decides how costs are distributed among customers.
If the Corp Comm has the backs for residential ratepayers and forces utilities to prove residents aren’t paying for the generating capacity for the data centers then rates shouldn’t rise because the center will bear those costs in their rates.
Kris Mayes calls for no new AI data centers as AZ faces water, power crunch
Sadly, the ACC has shown that they’re way more beholden to moneyed interests (and corporate campaign donors either indirectly or indirectly) than to actual voters and residential ratepayers, but that’s a conversation for another blog.
A Playbook for the Other 14 Counties
That’s why I’d encourage Arizona’s other 14 county health departments to spend some time with Pima County’s report.
You don’t need a data-center proposal sitting on the Board of Supervisors agenda before getting involved.
Use Pima’s framework. Look at your county’s baseline health and environmental data. Find vulnerable neighborhoods. Understand where the water and electricity would come from. Figure out what information you would need from a developer.
Then explain those issues to your supervisors before the big economic-development decisions get made.
Pima County didn’t conclude that data centers are good or bad for public health. They figured out what questions government should ask before saying yes.
