Data Center Chillers (3/3): How much can "free cooling" impactchiller service revenues? chillers work, the difference between air-cooled and water-cooled units (along with theireconomics), and run through"chiller gate"risk on equipment sales from both climatic andoperational riskperspectives.However,equipmentsales are onlyonepartofchillergate.Theotherquestion weoftenget asked is the impact reduced compressoruse could have onchiller service revenues. In this piece, we attempt to answer that question. +1 917344 8543varun.govindaraj@bernsteinsg.com Specialist Sales Steve Song+1917344 8401steve.song@bernsteinsg.com Before we can understand which parts of chiller service revenue are at risk, we first needto understand its constituent parts. There are four broad buckets.Not every installed chillergenerates allfour of these revenue streams, but we can assume representative averages.First,there is maintenancerevenue-largelypreventative anddeliveredthrough plannedservice agreements where technicians visit a unit at regular intervals. This could bemanagedthrough an upfrontpayment orannual contracts andtheduration can also varybefore renewal.Next youhave revenuefromequipmentmodernization/upgrades;these only happen once or twice over the life of a chiller(~25 years)and take place whena piece of technology is replaced (e.g., sensors added, compressorisupgraded).The lastmajorbucketisrepairrevenuefromequipmentoperational issues orbreakdowns.Repair revenue tends to increase in share as the unit gets older. In some cases there maybe ancillary revenue streams like software,butthese are a small share of the overallpicture (under 5%).We estimate that ~30% of the overall service revenue pool ismaintenancelabor,~20%islaborfromupgrades/modernization,and~15%islaborfromrepairs.~30%isrevenuefromparts (whichtendtoskewtowardsrepairandupgrades),with a final~5% comingfrom ancillary services (like software monitoring)Numbers will vary aroundthese averages,butthese are directionally accurate cuts. We believe maintenance andmodernization arenotat risk fromreducedusage of thechiller,whilerepair (bothlabor and parts)is and amountsto somewherebetween25-30%of thetotal revenue pool.This does notmean the entire repair revenuepool itself is atThis is also not likley to be an immediate headwind, the chiller needs to run for 5 - 10 yearsbeforethere's enough wear on the compressorto increase likelihood of breakdown.Weestimatea~1-2%headwindto theDC service revenue topline afterthat.Howmuch depends on the extent of free cooling that we see ;for example, chillers installed inNoVa will probably break down a lot less than those installed in Phoenix. But given we areseeing more chillers being setup in arid, desert environments whereland is plentiful thismay notbeasmuchofariskas somepeoplebelieve itwill be.TheOEMs themselves donotseem to be particularly worried for now, while some of them have told us they are modelingout the impact, the markettoday seems to be much morefocused on equipment sales vsservicerevenues. This is the final article in our three part series on chillers. You can view part one(discussing how chillers work,andthedifferent economics of air cooled and water cooledunits) and part two (discussing the risk of chiller gate on equipment sales through climaticandoperational risk analysis)usingthe hyperlinksprovided. We rate CARR Market-perform with a target price of $75.WerateJCIOutperformwithatargetpriceof $176.Werate VRTOutperform with a target priceof S416. MULTIEXHiBIT 1: Breakdown risk is concentrated in higher-use regimes; free-cooling reduces wear-and-tear and associated breakdown probability air-cooled and water-cooled units (along with their economics), and run through"chiller gate" risk from both climatic andoperational risk perspectives.However,equipmentrevenues are only onepart ofchillergate.The other side we oftenget askedabout isthe impactitcan have on chiller service revenues -driven byless intense use of the chiller (inparticular the compressor)in free cooling environments. First, let's clarifywhat wemeanbyfree-cooling.In certain situations,the ambienttemperature at the location a data centerhas been builtcan droptoapointwhere the chiller does not needto run atfull capacity (or run atall)for cooling outcomes tobeachieved.This canhappen inone of three ways.Thefirstisthroughair-side economizers,where coldair fromoutsidethedatacenterisdirectlypiped intocooltheroom.Whilethis workswithlowerrack densities,forcurrent Blackwell andhigherunitsit is notsufficient.Thenthereare water-sideeconomizers,whereeithercoldambientair or coldwaterfromacoolingtowercan cool thefacility water system (FwS)loop directly;this also reduces or eliminates the need for a chillerto make use of itscompressor.Finally,thereis refrigerant migrationwhere the condenserloopwater/air temperature is low enoughto condensetherefrigerant at lowpressures (so thecompressorcan run at reduced capacities or not at all). intensively,it sees less wearandhas a lower