that may have a direct effect on your bottom line. The question

VOLUME THREE NUMBER THREE SYNOPSIS A N H V A C N E W S L E T T E R F O R B U I L D I N G O W N E R S A N D M A N A G E R S In this issue... Cost of Owning a Chiller What is your bottom line?....... 1 What Should You Consider? Understanding chiller energy consumption................ 2 What is Chiller System Optimizer? Calculating the system part load... 3 Using Chiller System Optimizer to Help.......... 3 Optimization Can Save Time and Money...........4 Glossary of Terms...........5 Want to know more about system part load value? Call 1.800.CARRIER and request the publication HVAC Analysis, Vol. 3, No. 3, written specifically for consulting engineers, or visit our website at www.carrier.com System Part Load Value: A Case for Chiller System Optimization What Does It Cost to Own a Chiller? What does it cost to own two or three or four chillers installed in a central plant? The answer may not be what you think. Within the engineering community, there is an ongoing, behind-thescenes debate that may have a direct effect on your bottom line. The question centers on water chillers: how they are specified, and how they perform. The Air Conditioning and Refrigeration Institute (ARI) 1 develops and publishes technical standards that establish rating criteria and procedures for certifying equipment performance. 1 ARI is a national trade association representing manufacturers of more than 90% of the U.S. produced central air conditioning and commercial refrigeration equipment. In 1998, ARI revised their Standard 550/590, which defines the industry-accepted method of rating chiller performance. Since most chillers rarely operate at full load, the important criterion is how chillers perform at part load. The basic performance measure defined by ARI 550/590-98 is the Integrated Part Load Value (IPLV). IPLV is a single number representation of part load performance calculated at standard operating conditions. For conditions other than standard, ARI 550/590-98 also defines Non-Standard Part Load Value (NPLV) (see Glossary of Terms). The intent of IPLV (or NPLV) is to allow simple, uniform comparison of different chillers much the same as

Seasonal Energy Efficiency Ratio (SEER) is used for residential-type equipment. For instance, if choosing between two chillers, one with an IPLV of 0.50 kilowatts per ton, and a second with an IPLV of 0.54 kilowatts per ton, the apparent indication is that the first chiller is more efficient because it uses less electricity per ton of cooling. Unfortunately, this is not necessarily the case. Chilled water plants are complicated systems with different equipment types (i.e. chillers, pumps and cooling towers) that have interdependent performance relationships. These performance relationships are influenced by weather conditions, the building load profile, operating hours, the total number of chillers and other factors. In most cases then, real differences in energy consumption will not be what is indicated by the IPLV rating, because the IPLV specification used to buy the chiller is not representative of any particular project. It is based on only one chiller operating, average weather data, and an assumed building load profile. Indeed, ARI recommends that it is best to perform a comprehensive analysis that reflects the actual weather data, building load characteristics, number of chillers, operating hours and other factors unique to that specific project. What Is the Debate? When engineers specify electric water chillers, they define size, operating conditions, materials of construction, and the chiller s minimum efficiency. Efficiency has traditionally been expressed as maximum allowable kilowatts of energy consumption per ton of cooling (kw/ton). But, efficiency is elusive. Actual chiller efficiency will vary depending on a host of variables: whether the unit is operating 2 75% Energy and Maintenance SYNOPSIS at full load, minimum load or somewhere in between, the weather, and water temperatures. So, the great debate is how to define the minimum chiller performance. Using current standards, engineers specify chillers using IPLV, NPLV or full-load kw per ton. Some may even still use APLV 2. IPLV and NPLV are, at best, coarse indicators of anticipated performance at certain specific conditions. Full-load kw per ton is a meaningful performance indicator only for chillers that operate continuously at full load. Most chillers operate at part load more than 99% of the time. None of these performance gauges will assure you, as a Lifetime Cost of Chiller Ownership building owner, that you are buying chillers to best fit your needs. 24% Chiller and Installation Based on a new chiller installation. Installation includes the cost of piping, valves, insulation, pipe supports, electrical wiring and chiller foundation. 2 Application Part Load Value. APLV was eliminated from ARI Standard 550/590 in 1998. FIGURE 1 1% Engineering Design Why Should I Worry? If you are focused simply on minimizing first cost, it doesn t really matter. Have the engineer specify a chiller that seems reasonable on energy use, that fits the budget, and be done with it. But for most owners that s not enough. Today, most building owners and corporations are sophisticated in their approach to energy issues. Many have full-time energy managers and have established minimum energy standards to govern new construction. Every new building project has a construction budget and an operating budget; the informed owners recognize the need to consider economic trade-offs between both. Over the life of a chiller, energy and maintenance costs can total more than three to five times the original

cost to buy and install the chiller. Figure 1 shows a typical distribution of lifetime chiller costs for a new chiller installation. For a basic chilled water plant with two 500-ton centrifugal water chillers, the total lifetime cost can be about $1.8 million! Think about it. If you are preparing to make an asset purchase that will ultimately cost millions of dollars, do you want to use the same standardized single-digit decision method that a homeowner would use to buy a residential air conditioning unit? Probably not. Do you expect accurate, informed consultation from your engineer? You should. What Should My Engineer Do? The footnotes in ARI 550/590-1998 recommend performing a comprehensive analysis using actual weather data, building load characteristics, the number of chillers, actual operating hours, economizer capabilities, and the energy drawn by auxiliaries such as pumps and cooling towers. This is what you, as a building owner and a client, should require from your engineer. But, understand that a comprehensive analysis can be a lot of work. Many engineers, squeezed between schedules, architects and price competition, have not been able to get the fees it takes to do that much work. It is hard enough just getting the systems designed, drawn and specified in the time allotted, let alone perform a detailed, applicationspecific part-load analysis. There needs to be an inexpensive tool to significantly improve your decision-making information that can be made available with only marginal increase in engineering cost. That is why Carrier developed Chiller SystemOptimizer. What is Chiller System Optimizer? Chiller System Optimizer is a non-manufacturer specific software program that simplifies chilled water system analysis. Your engineer, working together with your local Carrier representative, can use the software to evaluate chiller performance by taking into consideration all of the system and location factors recommended by ARI. One of the program s unique features is that it calculates the System Part Load Value () as an alternative to ARI s standard IPLV and NPLV indicators. Like IPLV and NPLV, is a single numerical indicator of part-load chiller performance. The crucial difference is that it is calculated specifically for your project. This is important. is an accurate, representative indication of chiller performance operating under conditions unique to your project. So, How Does Chiller System Optimizer and Help? There is a misconception that chillers operate most efficiently at full load. People often then tend to think of a chiller s overall efficiency in terms of the full-load kw per ton. The truth is that a chiller can operate most efficiently wherever you want it to operate most efficiently. Manufacturers have at their FIGURE 2 Cooling Ton-hrs Per Year 175000 150000 125000 100000 75000 50000 25000 disposal huge assortments of compressor and heat exchanger combinations. For a given size of chiller, a manufacturer may have upwards of 10,000 different combinations of compressors, impellers, heat exchanger shells and tubes. If desired, a chiller can be selected to be most efficient at full load, 50% load, or anywhere in between. To make the best selection, it is necessary to first know where the most efficient point should be. That is a function of building load profile, hours of operation, the number of chillers operating at any given time, and the energy consumption of pumps, cooling towers and other auxiliaries. Figure 2 shows a typical building load profile overlaid with a plot of the system chiller profile. It can be seen Typical System Performance Profile Taken from Chiller System Optimizer Output Cooling Tons hrs 0 0 10 20 30 40 50 60 70 80 90 100 Building Load % Chiller System kw/ton 0.65 0.60 0.55 0.50 0.45 Chiller System kw/ton 3

that most of the time the building load is between 46 and 84% of full load. Conversely, the chiller efficiency varies, reaching its best efficiency during periods when the building load is at about 39% of full load. The weighted average operating point for both chillers is at about 0.483 kw/ton. This is the. Chiller System Optimizer provides the project criteria necessary to specify chillers that are optimized for your specific project. When you receive bids from chiller manufacturers based on these criteria, the quoted figures will be accurate, dependable numbers. Your engineer can input the bids and quoted performance data into Chiller System Optimizer to reliably evaluate life cycle cost. In addition to the chillers, Chiller System Optimizer considers the energy use of pumps, cooling towers and other system auxiliaries. This allows the engineer to evaluate and select the best chillers, and design a truly optimized system assuring the lowest life cycle cost to you. What Does The Tell Me? In the example in Figure 2, the chillers modeled have a full-load energy consumption of kw per ton and an NPLV of kw/ton. In this system and location, the chiller s actual part-load energy consumption, the, is about 0.483 kw per ton. This is over 19% better than the full-load measure and nearly 10% better than the NPLV measure. TABLE 1 BOSTON CHICAGO DALLAS MINNEAPOLIS ORLANDO SAN FANCISCO SEATTLE TUCSON Performance of a Typical Building System, Weather is the Only Factor Changed in Each City Minimum Maximum Variation Between Sites But Does It Really Make Any Difference? Competitor A (R123 Centrifugal) Chiller Full Load kw/ton Chiller NPLV We wondered the same thing, so we took an example project and made comparisons between eight different cities using the Chiller System Optimizer software. Table 1 shows the performance of a typical building system with a 1,000-ton load served by two 500-ton centrifugal chillers. The system is identical for each city shown: the same chillers, pumps, cooling towers and control sequence. Weather data was the only thing changed for each city. Also, two different makes of centrifugal chillers were compared in each city. The chillers were selected with identical full-load kw per ton ratings. To remain objective in this discussion, we did not include a Carrier chiller. 0.505 0.508 0.515 0.526 0.494 0.497 0.494 0.494 0.526 6.48% Competitor B (R134a Centrifugal) Chiller Full Load kw/ton Chiller NPLV 0.483 0.492 0.512 0.488 0.529 0.445 0.449 0.467 0.445 0.529 18.88% Variation Between Chillers -4.55% -3.25% -0.59% -3.69% 0.57% -11.01% -10.69% -5.78% The results are surprising. If you take a specification prepared for a project in Orlando and use it for a project in San Francisco, the actual performance may vary by nearly 19%. Also, depending upon which make of chiller is selected, the variation in performance can be as much as 11% at the same site even if both chillers meet the specification with identical full-load kw per ton ratings. continued on page 5 4 SYNOPSIS

5 continued from page 4 So, What Does That Mean to Me? Let s look at the most pronounced example, San Francisco. If the engineer s specification defines the minimum chiller performance as simply kw per ton at full load, all things being equal, the actual performance will vary by more than 11% depending on the make of chiller selected. Also, note that the Competitor B chiller with the larger NPLV () will actually operate more efficiently than the Competitor A machine () when evaluated as part of the overall system using sitespecific information. Why? Because NPLV (and IPLV) is calculated at a very narrowly defined profile that is not often representative of what is actually happening in the system. If the Competitor B chiller is installed, it will operate at about 0.445 kw/ton on average. This is nearly 17% better than the indicated NPLV and more than 25% better than the full-load rating. Is that what you expected? Is that what your engineer was expecting? Would you be surprised to know that, in spite of both chillers meeting the engineer s 3 Estimate of annual energy cost is based on electrical rates of $0.05 per kilowatt-hour and $7.50 per kilowatt demand, and includes the energy cost of pumps and cooling towers. specifications, if you buy the wrong chiller it will cost you more than $6,500 each year 3 in higher energy cost? Over the life of the chillers that adds up to an additional $163,000! Was that in the budget? Is it possible that you would like to be part of that life cycle decision? OK, But Doesn t Optimization Analysis Take a Lot of Time and Money? Not any more. In the old days, optimizing the chiller and system selections would take days of tedious calculations. Today, using Carrier s Chiller System Optimizer, the analysis to optimize chillers and system components for a specific project can be accomplished in a matter of hours. All your engineer needs to know is the peak load of your building and the load at one or two other points (e.g. the conditions where load equals zero). The program interpolates a load profile based on the actual anticipated facility operating hours. For special applications, the program allows you to define the actual building load profile. Other features: Weather is selected from actual weather data for 189 cities in the United States or 180 international locations. You can specify exact chiller performance from selections prepared by Carrier or other manufacturers. In addition, the program has performance templates of actual chillers from Carrier, Trane, York and McQuay. The computer model will also accept different combinations of chiller size, type (centrifugal, absorption, air-cooled, etc.) and manufacturer up to a total of 12 chillers, including chillers operating with variable speed drives. You can compare different chiller sequencing control options. Cooling tower performance can be approximated, or it can be accurately defined. Carrier has included a cooling tower performance algorithm in the software (provided by Baltimore Aircoil) to predict the actual tower performance. You are able to define specific economic criteria for your project: life cycle period, currency, different inflation rates for different items, energy rates for natural gas, electricity and steam, and the minimum rate of return. Costs for equipment, construction, maintenance, water treatment and other factors can be included. The program automatically performs the life cycle cost analysis between different options and systems that you define. When Should Chiller System Optimizer Be Used? Chiller System Optimizer fits into the design process in several locations. First, during preliminary design. When cooling loads have been approximated and the engineer is working with the architect to develop the building concept, Chiller System Optimizer can help determine the type of chilled water system to use. This allows the engineer to define space requirements for chillers, pumps and cooling towers. Second, during detailed design, Chiller System Optimizer can be used to define the exact criteria to be used

6 in the chiller specifications. Also, in retrofit projects, Chiller System Optimizer helps the engineer to specify chillers and equipment that will be truly optimized, providing the best return on investment for the project. Literally speaking, the time to use Chiller System Optimizer is now. Consider this: The time and expense of performing chilled water system optimization the old fashioned way has driven engineers into using simplified methods of specifying chilled water equipment. These methods are not indicative of true system performance, and leave you, the building owner, in the dark regarding operating cost. If you are uncertain about operating cost, then you are uncertain about 75% of the total cost of chiller ownership. Using Chiller System Optimizer, your engineer can prepare the meaningful analysis you need, simply and easily using inexpensive software that is not manufacturer specific. Chilled water systems are becoming more complex, and the options available to choose from are more varied and confusing than ever. The time for excuses is disappearing. Chiller System Optimizer is a tool available to your engineer through their local Carrier representative. APLV IPLV NPLV Application Part Load Value is a single number, part-load efficiency indicator* calculated using the ARI method referenced to selected conditions. APLV was introduced in ARI Standard 550-1988, but deleted from the 1998 edition. Integrated Part Load Value is a single number, part-load efficiency indicator* calculated using the ARI method at standard rating conditions. Introduced in ARI Standard 550-1986, the definition of IPLV was changed in ARI Standard 550/590-1998 to more closely reflect actual operating experience found in the field for a single chiller. Non-Standard Part Load Value is a single number, part-load efficiency indicator* calculated using the ARI method referenced to rating conditions other than ARI standard. The 1998 standard adopted NPLV for situations when a single chiller is not intended to operate at standard ARI rating conditions. Glossary of Terms System Part Load Value is a project-specific, single number, part-load efficiency indicator introduced by Carrier that is calculated using the equation form defined in ARI 550/590-1998. Unlike IPLV or NPLV, the factors used to calculate are project specific and consider multiple-chiller applications, actual operating hours, and project-specific operating conditions. * ARI Standard 550-88 uses the term figure of merit. The word indicator is used as a clarification. The use of a single number performance value is to provide an indication of equipment performance relative to other equipment at defined conditions.