new ari rating allows more accurate chiller-energy specification

7/22/2019 New ARI Rating Allows More Accurate Chiller-Energy Specification

1/4

An accurate rating method for chiller per-

formance is an essential tool for evaluatingand comparing equipment. Now there is anew rating developed by the Air-conditioningand Refrigeration Institute (ARI) that hasreplaced the previous Integrated Part LoadValue/Application Part Load Value(IPLV/APLV) rating. The new rating closelytracks real-world, chiller-energy perfor-mance by more accurately accounting forchiller operation at off-design conditions.

The new rating termed IPLV/NPLV forIntegrated Part Load Value/Non-standardPart Load Value is part of the new ARIStandard 550/590-98. The new standard,which becomes effective in December 1998,also includes a revision of the evaporatorfouling factor from 0.00025 to 0.0001.Starting immediately, all chiller manufactur-ers will begin certifying chiller performanceto the new standard.

The IPLV/NPLV rating is easy to use itcan be employed exactly the same way asthe old IPLV/APLV rating. In fact, to incorpo-rate the new standard into specifications,engineers can simply replace the termAPLV with the term NPLV, change theevaporator fouling factor to 0.0001, andreplace any references to ARI Standard550-92 or ARI Standard 590-92 with ARIStandard 550/590-98. No other change isrequired. The only change is in the waychiller manufacturers calculate the rating.

Why was the old rating changed?Introduction of the new IPLV/NPLV rating

solves several problems with the oldIPLV/APLV rating, which was based on sev-eral less-than-realistic assumptions. Mostnotably, the IPLV/APLV rating failed as anaccurate measure of chiller performance atoff-design, real-world operating conditions.The IPLV/NPLV rating now provides a morerealistic model of off-design performance.

Chillers rarely operate at design condi-tions, because design conditions mean thesimultaneousoccurrence of both designload and design Entering Condenser WaterTemperature (ECWT) or design Entering DryBulb (EDB) temperature. But design ECWTor EDB occur during less than 1% of chilleroperating hours. That means that over 99%of potential chiller operating hours are spentat off-design conditions: reduced loads,reduced ECWT/EDB, or both. TheIPLV/NPLV rating accounts for this moreaccurately.

How has the rating changed?

To understand the new rating, it is neces-sary to look at the factors involved in the oldIPLV/APLV rating.

As it stood, the old IPLV/APLV rating wasstill a much better indicator of chiller perfor-mance than the design kW/Ton rating,because it took into account some off-designconditions. Unfortunately, the old rating wasbased on limited assumptions that werentaccurate for the operation of most buildings.

Old assumptions

Atlanta weather data was used to repre-sent the average weather for the entire

U.S. The result was that the rating wassomewhat skewed toward warmer andmore humid weather.

The old rating improperly representeddesign conditions as occurring more often

New ARI rating allowsmore accurate chiller-energy specification


2/4

than they do in the real world. Specifically,the old rating said 17% of chiller operatinghours were spent at design conditions of100% load and 85F ECWT (for water-cooled chillers) or 95F EDB (for air-cooledchillers). In reality, these conditions occursimultaneously during less than 1% ofchiller operating hours.

The old rating used an antiquated rule-of-thumb that assumed that for every 10%reduction in load, ECWT declined 2.5Fand EDB declined 4F. However, in thereal world, the ECWT decline averages4F and the EDB decline averages 6F per10% load reduction.

The old rating assumed all buildings hadan airside economizer and operated 12hours/day, 5 days/week. However, a 1992DOE study (combined with a BOMA study)found that only 32% of commercial build-ings have both of these operational char-acteristics. The result was the old rating

underestimated chiller operating hours forover two-thirds of the buildings.

New assumptions

By avoiding these misleading assump-tions, the new IPLV/NPLV rating overcomesthe limitations of the old IPLV/APLV rating. The new rating uses national weather data,

based on a weighted average of the 29cities where 80% of the chillers were pur-chased over a 25-year period, as reportedby ARI.

The new rating recognizes that most oper-ating hours are spent at off-design condi-

tions. The new rating factors in the effect of cold-er ECWT and colder EDB.

The new rating uses weighted averages offour different building operating scenarios:- 24 hours/day, 7 days/week, without an

airside economizer- 24 hours/day, 7 days/week, with an air-

side economizer- 12 hours/day, 5 days/week, without an

airside economizer- 12 hours/day, 5 days/week, with an air-

side economizer

Comparison of old and new ratings

Each rating is a blending of the kW/Ton atfour rating points: 100%, 75%, 50%, and25% load. The formulas used to calculatethe ratings are a bit complex, which makescomparison of the old and new ratings diffi-cult. It is more instructive to compare thefour points that make up each rating. A com-

parison of the old and new rating points inTable 1 illustrates the two major changes.

First, the new rating assumes that only1% of chiller operating hours are spent atdesign conditions, instead of the 17%assumed by the old rating. Second, the newrating assumes that ECWT and EDB declinemore rapidly than the old rating did.

How closely does the new rating predictperformance?

Using these new assumptions, howclosely does the new IPLV/NPLV rating trackchiller performance? To find out, lets do a

detailed energy analysis of an examplechiller, and then compare that performanceto both the old and new IPLV ratings.

Well start by using weather data fromAtlanta, Dallas, Baltimore, Chicago, andNew York City. A load line, with internalloads averaging about 40% of the total load,was then superimposed on this weatherdata. For chiller efficiency, part-load curvesfrom the major chiller manufacturers wereaveraged together to create the genericcentrifugal chiller. Weighted averagekW/Ton values were then calculated for thefour different building operating scenarios.

The results are shown in Figure 1.A quick perusal of Figure 1 clearly indi-cates the improved accuracy of the newIPLV/NPLV rating. It is equally accurate forscrew and reciprocating chillers.

What are the benefits of the new rating?With the new IPLV/NPLV rating, engi-

neers now have a more accurate and con-venient specification tool. It is an easy wayto gauge actual chiller performance,because the new rating accounts for morerealistic operating conditions. In addition,only a detailed energy analysis could pro-vide a more comprehensive indication of

actual performance.The new rating will allow owners to make

more informed capital-investment decisions,because it will provide a more realistic esti-mate of actual chiller operating costs. Plus,it will help determine more accurate pay-backs in both new-construction and replace-ment projects. The old rating would add asmuch as 25% more time to the chiller-invest-ment payback period. This often influenceddecision-makers to choose less efficientchillers in new construction projects, ordelay the replacement of inefficient chillersin renovation projects.

Table 1: Comparison of Old and New Rating Points

Rating Load Weighting ECWT EDBPoint (%) (%) (F) (F)

Old New Old New Old New

A 100 17 1 85.00 85.00 95.00 95.00

B 75 39 42 78.75 75.00 85.00 80.00

C 50 33 45 72.50 65.00 75.00 65.00

D 25 11 12 66.25 65.001 65.00 55.002

Notes:1. Would have been 55F, but industry testing facilities incapable of testing that low.2. Would have been 50F, but industry testing facilities incapable of testing that low.


3/4

If we put a chiller on a test block, wherewe have control of both the load and theECWT/EDB, we can see why this is so. IfECWT/EDB is held constant while the loadis reduced, a chiller sees relatively littlechange in kW/Ton usually only 5 to 6%.However, if the load is held constant whilethe ECWT/EDB is reduced, kW/Ton can fall

by 45 to 50%! The use of multiple chillers ina plant only affects the load each chillersees, not the ECWT/EDB. So multiplechillers have very little effect on chiller-plantaverage kW/Ton.

Table 2 shows an example comparing theaverage plant kW/Ton for chiller plants usingone through six chillers to handle the samecooling duty. The kW/Ton at design condi-tions for each chiller was 0.583, and thenew IPLV was 0.488.

Clearly, the kW/Ton averages from 0.482to 0.492 are not only close to one another,but also close to the new IPLV rating of

0.488 (by the way, the old IPLV would havebeen a misleading 0.522). This demon-strates that the rating can be used to predictchiller performance in multiple-chiller sys-tems, as well as single-chiller systems.

What should be done now?In addition to substituting NPLV for

APLV, 0.0001 for 0.00025 and550/590-98 for both 550-92 and 590-92in their specifications, engineers can noweliminate the specification of design kW/Ton.A design kW/Ton rating is merely an efficien-cy at one condition a condition that occurs

during less than 1% of chiller operatinghours. It provides no indication of off-design performance, which can vary wide-ly from chiller to chiller. In fact, chillers withthe best design kW/Ton may have the worstIPLV/NPLV performance because they wereoptimized for design conditions.

Its true that design kW can affectdemand costs, but this only becomes signifi-

The environmental impact of such chiller-replacement decisions can be staggering. Ifonly 10% more of the existing, inefficient,CFC chillers were replaced with more effi-cient units, global warming could bereduced by a half-million tons of CO2 over afive-year period. The phase-out of CFCrefrigerants would also be accelerated.

What about multiple-chiller plants?Appendix D of the new standard states

that the new rating was developed forsingle-chiller plants, and is not representa-

tive of multiple-chiller plants. It further statesthat individual chillers in a multiple-chillerplant spend, on average, more operatinghours at higher loads than would a singlechiller handling the same duty.

What the Appendix doesnt say is that theinvestigation of multiple-chiller plants wasnever undertaken. So, it was unknownwhether the new rating was accurate formultiple-chiller plants.

That investigation has now been conduct-ed by YORK, one of the developers of thenew rating. The investigation found that thenew rating accurately tracks the perfor-

mance of multiple-chiller plants, as well assingle-chiller plants.

The explanation is rather simple. It is truethat individual chillers in a multiple-chillerplant spend, on average, more operatinghours at higher loads than would a singlechiller handling the same duty. However, itis also true that those additional high-loadoperating hours are spent at lowerECWT/EDB. The result is that chillers in amultiple-chiller plant use the same or evenless energy than a single chiller handling thesame duty.

Table 2: Average Chiller-Plant kW/Ton vs. New IPLV of 0.488

Operating Number of Chillers in PlantScenario1 One Two Three Four Five Six

1 0.496 0.476 0.475 0.479 0.481 0.4822 0.474 0.485 0.484 0.492 0.492 0.496

3 0.494 0.489 0.481 0.486 0.488 0.490

4 0.487 0.481 0.489 0.496 0.496 0.501

Avg. 0.488 0.483 0.482 0.488 0.489 0.492

IPLV% +0.4 -1.0 -1.2 +0.2 +0.2 +0.8

Notes:1. Scenario 1 = 24 hours/day, 7 days/week operation, without airside economizer

Scenario 2 = 24 hours/day, 7 days/week operation, with airside economizer

Scenario 3 = 12 hours/day, 5 days/week operation, without airside economizer

Scenario 4 = 12 hours/day, 5 days/week operation, with airside economizer

Atlanta Baltimore ChicagoDallas New York Old IPLV New IPLV

0.54

0.53

0.52

0.51

0.50

0.49

kW/Ton

1 2

Operating Scenario

3 4

Figure 1. New and Old IPLV Ratings asPredictors of Chiller Performance


4/4

A new standard for a new eraThe introduction of the new IPLV/NPLV

rating is a tremendous development forengineers and owners alike. It is the bestbasis to date for writing a chiller-efficiencyspecification. Nothing short of a detailedenergy analysis is better at determiningchiller performance for both single- and

multiple-chiller plants.With the IPLV/NPLV rating, engineers can

easily produce a more accurate specifica-tion. Simply use the IPLV/NPLV number asprovided by the manufacturer.

The IPLV/NPLV rating also helps ownersascertain the best chiller for their applica-tion. Comparisons of chiller performancecan be made quickly and confidently. Usingthe rating also makes it easy to projectannual energy costs using this formula:

Annual Energy Costs = NPLV x $/kWh xAverage Chiller Load x Operating Hours

For more information about the newIPLV/NPLV rating, contact your local YORKoffice.

cant with ratcheted rate structures. Ratchetsare likely to disappear, and demand costsbecome less important, with the advent ofelectric-utility deregulation and real-timepricing.

In addition, the thermal flywheel of thebuilding moves the peak load on the chillersback a few hours later than the peak kW

draw for the building itself. The maximumdaily ECWT/EDB also occurs in the laterafternoon. As a result, the chillers usuallycontribute only 80 to 90% of their monthlypeaks to the actual building peaks.

Instead of specifying design kW/Ton, thespecification should be written so the chillermanufacturer isnt restricted from biddingequal efficiency. Doing so allows the man-ufacturer to meet the specified IPLV/NPLVrating, even if the chiller has a differentdesign kW/Ton.

Nevertheless, it is still necessary to speci-fy maximum full-load amps for wire sizing.

Thats because current requirements atdesign conditions will dictate the wiring sizerequired to handle the load. Thus, full-loadamps should be in the specification, but itshould be specified as a maximum thesame way overall equipment dimensions arespecified so that all manufacturers can fitthe space.

Proud Sponsorof the 2000U.S. Olympic Team

36USC380

P.O. Box 1592, York, Pennsylvania USA 17405-1592Copyright by YORK International Corporation 1998

Subject to change without notice. Printed in USA.ALL RIGHTS RESERVED

new ari rating allows more accurate chiller-energy specification

Documents