PUE versus WUE in a Nordic climate — a tradeoff file, not a winner

Reading order

Start with the two names. PUE is energy over energy. WUE is water over energy. The Green Grid’s PUE examination and ISO/IEC 30134-2 already in this register own the energy ratio. The 2011 WUE paper and ISO/IEC 30134-9 already own the water ratio. Do not let a slide treat them as spellings of one score. A hall can move one cell without filling the other.

Then write the cooling object that would move both cells. Outdoor air, a seawater loop, an evaporative or adiabatic rejector, a compressor plant and a liquid indoor loop are different machines. The IEA water box is useful because it refuses one thirst number: water varies with cooling technology, climate and the electricity source; about two-thirds of 2023 global consumption sat with energy supply and about one quarter with direct cooling. Direct-expansion, the same box says, is many times less water-intensive than airside economisers and adiabatic systems with water-cooled chillers. That sentence is the tradeoff the slide usually hides. A design that chases a lower PUE with evaporative rejection can raise site water. A design that avoids on-site water can still have a source-water cell in the electricity.

Only then open the Swedish operator rows and the weather row. Meta Luleå publishes outdoor air, hydropower and PUE 1.09. atNorth SWE01 publishes liquid cooling, a design PUE of 1.2 and heat recovery with Stockholm Exergi. EcoDataCenter Falun and GleSYS Falkenberg publish 1.2. Digital Realty STO1 publishes remote seawater and “cooling-free” operation for over half the year. SMHI publishes station temperatures. The water-use guide already owns how to read a WUE claim. The standards guide already owns KPI names. The energy-efficiency guide already owns PUE comparability. This page does not steal those matrices. It only asks whether a Nordic winter is being used to fill a missing cell.

• Name PUE and WUE as two ratios before the climate slogan.
• Name the rejector that would move water, energy, or both.
• Keep IEA global litres off a Luleå or Kista cell.
• Parent water-use, standards and efficiency pages last.

What the tradeoff actually is

The tradeoff is a design choice about where heat leaves the hall, not a Nordic medal. Compressors use electricity and can lift PUE. Evaporative or adiabatic rejection can cut compressor hours and raise site water. Outdoor air can cut both compressor hours and evaporative litres in a cold, dry hour, and still leave humidification, makeup or a source-water cell. Seawater moves the water object to a once-through or remote loop. The IEA box is global. It does not name Luleå, Kista or Falkenberg.

PUE cannot report that choice alone. A published PUE of 1.09 or a design PUE of 1.2 is an energy ratio with whatever boundary the operator used. It is not a litre. WUE cannot report the energy choice alone. A missing Swedish WUE is a missing cell, not a zero. The Union reporting rule asks for both families at the 500 kilowatt IT-power threshold. A reporting duty is not a measured Falun year and not a licence to treat one ratio as the other.

A Nordic winter is a climate series, not a permit. SMHI’s station temperatures and 30-year normals describe weather. They do not sign a cooling plant. Meta’s outdoor-air sentence is one northern operator plant from 2013. Digital Realty’s “cooling-free” wording is one Kista seawater sentence. Neither page publishes the other metric. atNorth’s liquid-cooling and heat-recovery sentence can change the rejector and the heat-buyer file. It does not print a WUE.

Source water stays a third cell. The Green Grid’s WUEsource adds water used to make the electricity. Sweden’s generation mix is a national energy story. It does not fill a site WUE, and it does not travel as a litre permit to Staffanstorp. Bidding areas SE1 to SE4 do not allocate water.

What named primary sources show

The Green Grid PUE paper and ISO/IEC 30134-2 show the energy ratio and measurement categories. The 2011 WUE paper and ISO/IEC 30134-9 show the water ratio and the site/source split. Delegated Regulation (EU) 2024/1364 shows a reporting duty for both families at 500 kilowatts of installed IT power demand. The IEA demand chapter shows the 7 percent and more-than-30 percent cooling shares. The IEA water box shows 560 and 1,200 billion litres, the United States 100 megawatt example, the two-thirds / one-quarter split and the direct-expansion sentence.

Meta Luleå shows outdoor air and PUE 1.09. atNorth SWE01 shows liquid cooling and design PUE 1.2. EcoDataCenter Falun and GleSYS Falkenberg show 1.2. Digital Realty STO1 shows seawater and a cooling-free-hours sentence. SMHI shows weather. None of those operator rows is a Swedish WUE table.

A matrix instead of a Nordic winner

The table is a permission table. A filled cell is a claim the named URL can carry. An empty cell means this page will not invent the missing Swedish WUE or convert a winter photograph into both ratios.

Do not treat the IEA 7 percent cooling share as a Luleå PUE, and do not treat the United States litre example as a Kista WUE.

Common misreads

The first misread is to treat a low PUE as a low WUE. They are different ratios. The second is to treat a Nordic winter as a measured zero on both. Weather is not a year of litres and not a year of facility energy. The third is to treat the IEA sentence that direct-expansion uses many times less water than adiabatic systems as a ban on outdoor air. The sentence is a global technology comparison. It is not a Swedish permit.

The fourth is to treat Meta’s 1.09 or atNorth’s design 1.2 as a national league. Different years and boundaries stay on different rows. The fifth is to treat STO1’s cooling-free hours as a closed water file. Seawater is still a water object. The sixth is to fill a missing Swedish WUE from the United States 100 megawatt litre example.

What to ask next

If a brief says a Nordic hall wins both ratios, ask which named URL publishes a dated PUE and a dated WUE for that hall, with boundaries. If the URL only says outdoor air, seawater, design 1.2 or a SMHI normal, the brief has left the source.

Then open the water-use page, the energy-efficiency page and the standards page. Those pages already own WUE reading, PUE comparability and KPI names. This page only asks whether the winter is being used as a second metric.

1. Are PUE and WUE still two ratios?
2. Which rejector is named — outdoor air, seawater, evaporative, compressor, liquid?
3. Is the number design, measured or marketing?
4. Is an IEA global litre or cooling share being treated as a Swedish hall?
5. Has a SMHI normal been asked to close a water or energy cell?

What this does and does not prove

This page proves that PUE and WUE are sibling ratios with different numerators, that the IEA water box describes a technology tradeoff in global language, and that named Swedish operator pages publish PUE or cooling sentences without a paired WUE table.

It does not prove a Swedish national WUE. It does not prove that a Nordic winter fills both cells. It does not treat Meta’s 1.09 as a WUE or STO1’s seawater as a PUE.

Empty cells stay empty. Until a later primary source names a Swedish hall with both dated ratios and a boundary, the tradeoff stays a method, not a census.