APPLICATION NOTE An In-Depth Examination of an Energy Summary Efficiency Technology Thermal energy storage (TES) systems shift cooling energy use to non-peak times. They chill storage media such as water, ice, or a phase-change material Thermal Energy during periods of low cooling demand for use later to meet air-conditioning loads. Operating strategies are gener- Storage Strategies ally classified as either full storage or partial storage, referring to the amount for Commercial of cooling load transferred from on-peak HVAC Systems to off-peak. TES systems are applicable in most commercial and industrial facilities, but certain criteria must be met for eco- nomic feasibility. A system can be ap- propriate when maximum cooling load is significantly higher than average load. High demand charges, and a significant differential between on-peak and off- peak rates, also help make TES sys- tems economic. They may also be ap- Summary ...... 1 propriate where more chiller capacity is needed for an existing system, or where How This Technology back-up or redundant cooling capacity Saves Energy ...... 2 is desirable.

Types of Demand and Energy Besides shifting load, TES systems may Efficiency Measures ...... 3 also reduce energy consumption, de- pending on site-specific design, notably Applicability ...... 7 where chillers can be operated at full Field Observations to Assess load during the night. Also, pumping Feasibility...... 7 energy and fan energy can be reduced by lowering the temperature of the wa- Cost and Service Life ...... 11 ter, and therefore the air temperature, affecting the quantity of air circulation Laws, Codes, and Regulations...... 13 required. Definitions of Key Terms ...... 13 Capital costs tend to be higher than a References to More Information...... 14 conventional direct-cooling system, but other economic factors can reduce such Major Manufacturers ...... 14 costs. In new construction, ductwork

Copyright © May 1997, Pacific Gas and Electric Company, all rights reserved. Revised 4/25/97 could be smaller, allowing more usable does require the chiller to work harder space. Or a TES system may enable to cool the system down to the required reduction in electrical capacity, reducing lower temperatures (for ice storage); the cost of electrical service for a new or and energy is needed to pump fluids in expanding facility. and out of storage.

But a number of design options can How This Technology make TES systems more energy- efficient than nonstorage systems. Stor- Saves Energy age systems let chillers operate at full load all night , versus operating at full or In a TES system, a storage medium is part load during the day. chilled during periods of low cooling demand, and the stored cooling is used Depending on the system configuration, later to meet air-conditioning load or the chiller may be smaller than would be process cooling loads. required for direct cooling, allowing smaller auxiliaries such as cooling- The system consists of a storage me- tower fans, condenser water pumps, or dium in a tank, a packaged chiller or condenser fans. Pumping energy can built-up refrigeration system, and inter- be reduced by increasing the chilled connecting piping, pumps, and controls. water temperature range; fan energy The storage medium is generally water, can be cut with colder air distribution. ice, or a phase-change1 material Storage systems can also make in- (sometimes called a eutectic salt); it is creased use of heat recovery and wa- typically chilled to lower temperatures terside economizer strategies. than would be required for direct cooling to keep the storage tank size within economic limits. Figure 1 illustrates the basic operation of a system that uses Discharging chilled water.

Load shifting is typically the main rea- Building Load son to install a TES system. Cool stor- Distribution age systems can significantly cut oper- Pump Warm ating costs by cooling with cheaper off- Storage peak energy, and reducing or eliminat- ing on-peak demand charges. Cool

These systems have a reputation for consuming more energy than nonstor- Primary Pump age systems. This has often been true Chiller where demand reduction was the pri- Charging mary design objective. Cool storage

1 Bold italicized words are defined in the section Figure 1: Thermal Energy Storage title “Definition of Key Terms.” System with Stratified Chilled Water Storage (Source: ASHRAE)*

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 2 Tank volume is affected by the separa- Types of Demand and tion maintained between the stored cold Energy Efficiency Measures water and the warm return water. Natu- ral stratification has emerged as the preferred approach, because of its low TES systems can be characterized by cost and superior performance. Colder storage medium and storage technol- water remains at the bottom and ogy. Storage media include chilled wa- warmer, lighter water remains at the top. ter, ice, and phase-change materials, Specially designed diffusers transfer which differ in their operating charac- water into and out of a storage tank at a teristics and physical requirements for low velocity to minimize mixing. storing energy. Storage technologies include chilled water tanks, ice systems, The figure of merit (FOM) is a measure and phase-change materials. of a tank’s ability to maintain such sepa- ration; it indicates the effective percent- Chilled Water Storage age of the total volume that will be available to provide usable cooling. These systems use the sensible heat Well-designed stratified tanks typically capacity of water (1 Btu per pound per have FOMs of 85 to 95 percent. degree Fahrenheit) to store cooling. Tank volume depends on the tempera- The practical minimum storage volume ture difference between the water sup- for chilled water is approximately 10.7 plied from storage and the water re- cubic feet per ton-hour at a 20°F tem- turning from the load, and the degree of perature difference. separation between warm and cold wa- ter in the storage tank. Where most Chilled Water System Example conventional nonstorage HVAC systems Application operate on temperature differentials of 10° to 12°F, chilled water systems gen- Needing an additional 3,750 tons of erally need a differential of at least 16°F peak cooling capacity for a 250,000- to keep the storage tank size reason- square-foot addition to its Dallas head- able. A difference of 20°F is the practi- quarters, Texas Instruments chose a cal maximum for most building cooling chilled water thermal storage system: applications, although a few systems Adding chiller capacity would have cost exceed 30°F. about $10 million, while the TES system cost about $7 million—and a $200/kW Chilled water is generally stored at 39°F utility rebate reduced this to about $5.75 to 42°F, temperatures directly compati- million. The new system also has cut ble with most conventional water chillers operating costs about $1.5 million per and distribution systems. Return tem- year and allowed postponement of an peratures of 58° to 60°F or higher are expansion of the facility’s high-voltage desirable to maximize the tank tem- substation. perature difference and minimize tank volume. The TES system uses a 5.2-million- gallon, thermally stratified chilled water storage tank, built under a parking lot.

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 3 This pre-stressed, cylindrical concrete reservoir, 140 feet in diameter and 45 feet in height, has a design discharge rate of 7,500 tons—delivering 40°F wa- Building Ice Harvester ter at 12,000 gallons per minute (gpm) Load Chiller which returns to the tank at 55°F.

This system shifts 5.1 MW (35 percent) of existing electric chiller load to off- peak hours—about 8.5 percent of total facility demand. Five existing chillers Chilled Ice-water Ice / Water with total capacity of 6,500 tons now Water Pump Mixture remain off during the day. Mechanical Pump cooling systems now operate more effi- ciently because they are more fully Figure 2: Ice Harvesting loaded (0.85 kW/ton versus the previ- (Source: ASHRAE)* ous 0.95 kW/ton) and produce 13 per- cent more annual ton-hours of cooling. a higher temperature rise at the load, up Ice Storage to 25°F. The following technologies are used: Ice thermal storage uses the latent heat of fusion of water (144 Btu per pound). · Ice harvesting. Ice is formed on an Storage volume is generally in the evaporator surface and periodically re- range of 2.4 to 3.3 cubic feet per ton- leased into a tank partially filled with hour, depending on the specific ice- water. Cold water is pumped from the storage technology. tank to meet the cooling load. Return water is then pumped over ice in the Thermal energy is stored in ice at 32°F, tank. Refer to Figure 2. the freezing point of water. The equip- ment must provide charging fluid at · External melt ice-on-coil. Ice is temperatures of 15° to 26°F, below the formed on submerged pipes or tubes normal operating range of conventional through which a refrigerant or secon- cooling equipment for air-conditioning. dary fluid is circulated. Storage is dis- Depending on the storage technology, charged by circulating the water that special ice-making equipment is used or surrounds the pipes, melting the ice standard chillers are selected for low- from the outside. temperature duty. The heat transfer fluid may be the refrigerant itself or a secon- · Internal melt ice-on-coil. Ice is dary coolant such as glycol with water formed on submerged pipes or tubes, or some other antifreeze solution. as in the external melt system. Cooling is discharged by circulating warm cool- The low temperature of ice can also ant through the pipes, melting the ice provide lower temperature air for cool- from the inside. ing. The lower-temperature chilled water supply available from ice storage allows

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 4 · Encapsulated ice. Water inside selected temperatures. Most common is submerged plastic containers freezes a mixture that stores 41 Btu per pound and thaws as cold or warm coolant is at its melting/freezing point of 47°F. circulated through the tank holding the This material is encapsulated in rectan- containers. gular plastic containers, which are stacked in a storage tank through which · Ice slurry. Water in a water/glycol water is circulated. The net storage vol- solution is frozen into a slurry like the ume of such a system is approximately ice in a “Sno-Cone” and pumped to a six cubic feet per ton-hour. storage tank. Slurries can be pumped from the tank to heat exchangers or di- The 47°F phase-change point of this rectly to cooling coils, resulting in high material allows the use of standard energy transport rates. chilling equipment. Discharge tempera- tures are higher than the supply tem- The most common commercial technol- peratures of most conventional cooling ogy today is internal melt ice-on-coil. systems, so operating strategies may be External melt and ice-harvesting sys- limited. tems are more common in industrial ap- plications, and can also be applied in Phase-change materials are also avail- commercial buildings. Encapsulated ice able for lowering the storage tempera- systems are also suitable for many tures of ice systems. Additives on the commercial applications. market reduce freezing temperatures to 28° and 12°F in ice storage tanks; they Ice Storage Example Application reduce the latent heat capacity of water, as well as lower the freezing point. The The Seafirst Building in Bellevue, material is highly corrosive, so care Washington uses an ice storage system must be used in applying it. and cold air distribution with smaller- than-typical ducting. An Electric Power Operating and Control Research Institute comparison showed Strategies that this increased gross construction costs but reduced construction cost per TES operating strategies are generally square foot by about $4, because re- classified as either full storage or par- duced floor-to-floor heights allowed 21 tial storage, referring to the amount of stories within a height that would nor- cooling load transferred from on-peak mally accommodate only 20—thus periods. Strategies for operation at less adding 13,000 square feet of rentable than design loads include chiller priority space. Smaller mechanical rooms and storage priority control. added another 4,000 square feet. This provides about $340,000 per year of The period during which a system must additional income. reduce electric demand is generally called “on-peak,” often but not neces- Phase-Change Material Storage sarily synonymous with on-peak hours defined by the electric utility. In some Phase-change materials, or eutectic facilities the period may actually be salts, are available to melt and freeze at shorter than the utility on-peak period.

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 5 Peak electric demand from cooling also Load Chiller On may not occur simultaneously with the Chiller Charging Storage peak facility demand. Chiller Meets Load Directly Storage Meets Load Cool storage systems are usually sized to generate enough cooling in 24 hours to meet all the loads occurring during that period, but some applications use Tons longer cycles.

Full Storage 24-hour Period Full-storage, or load-shifting (shown in Figure 3), shifts the entire on-peak Figure 4: Partial-storage Load- cooling load to off-peak hours and usu- Leveling Operating Strategy ally operates at full capacity to charge (Source: ASHRAE)* storage during all non-peak hours. On- meets the rest. Such operating strate- Load Chiller On gies can be further subdivided into load- Chiller Charging Storage leveling and demand-limiting, Figures 4 Chiller Meets Load Directly and 5. Storage Meets Load In a load-leveling system, the chiller typically runs at full capacity for 24 hours on the design day. When the load

Tons is less than the chiller output, the ex- cess charges storage. When the load exceeds chiller capacity, the additional requirement is discharged from storage. 24-hour Period

Load Chiller On Figure 3: Full Storage Operating Chiller Charging Storage Strategy (Source: ASHRAE)* Chiller Meets Load Directly Storage Meets Load peak, all cooling loads are met from storage, and the chiller does not run. A full-storage system requires relatively Reduced On-peak large chiller and storage capacities and Demand is most attractive where on-peak de- Tons mand charges are high or the on-peak period is short. 24-hour Period Partial Storage

With this strategy chiller capacity is less Figure 5: Partial-storage Demand- Limiting Operating Strategy than design load. The chiller meets part (Source: ASHRAE)* of the on-peak cooling load and storage

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 6 This approach minimizes required chiller and storage capacities and is Applicability particularly attractive where peak cool- ing load is much higher than average Thermal energy storage can be used in load. In many systems, cooling loads virtually any building; the merits are during off-design periods are small compelling in the right situations. The enough that partial-storage systems main issues are the type of storage may be operated as full-storage sys- system and the amount of cooling load tems. This can increase savings but to be shifted. Before embarking on a care must be taken to not deplete stor- cool storage project, however, one age before the on-peak period is over. should consider the field observation guidelines below. In addition, alternative A demand-limiting partial-storage sys- approaches should be considered for tem operates the chiller at reduced ca- each project. pacity on-peak. The chiller may be con- trolled to limit the facility demand at the billing meter. This strategy falls between Field Observations to load shifting and load leveling. Demand savings and equipment costs are higher Assess Feasibility than for load-leveling, and lower than for load-shifting. Here too care must be This section discusses observations taken not to deplete storage before the and checks that can ensure a TES sys- end of the on-peak period. tem is appropriate and is installed and working properly. Additional variations on the full- and partial-storage strategies are possible Related to Applicability by scheduling the operation of multiple chillers. The many successful TES systems op- erating today demonstrate that the Partial-storage systems use one of two technology can provide significant control strategies to divide the load benefits. However, many cool storage between chiller and storage. A chiller- systems have failed to perform as pre- priority strategy uses the chiller to di- dicted because they did not meet the rectly meet as much of the load as pos- criteria for applicability cited below. sible. Cooling is supplied from storage only when load exceeds chiller capacity. Cool storage systems are most suitable Storage-priority meets as much of the where any of the following criteria apply: load as possible from stored cooling, using the chiller only when daily load · The maximum cooling load of the exceeds total stored cooling capacity. facility is significantly higher than the Some systems use combinations of average load. The higher the ratio of these strategies. For example, chiller peak load to average load, the greater priority during off-peak daytime hours, the potential. and storage priority during on-peak hours.

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 7 · The electric utility rate structure · The maximum cooling load of the includes high demand charges, facility is very close to the average ratchet charges, or a high differential load. A TES system would offer little between on- and off-peak energy rates. opportunity to downsize chilling equip- The economics are particularly attrac- ment. tive where the cost of on-peak demand and energy is high. · On-peak demand charges are low and there is little or no difference be- · An existing cooling system is be- tween the costs of on- and off-peak en- ing expanded. The cost of adding cool ergy. There is little economic value for storage capacity can be much less than customers to shift cooling to off-peak the cost of adding new chillers. periods.

· An existing tank suitable for cool · The space available for storage is storage use is available. In some ret- limited, there is no space available, the rofits, particularly in industrial applica- cost of making the space available is tions, using existing tanks can reduce high, or the value of the space for some the cost of installing cool storage. other use is high.

· Electric power available at the site · The cooling load is too small to is limited. Where expensive trans- justify the expense of a storage sys- formers or switchgear would otherwise tem. Typically, a peak load of 100 tons have to be added, the reduction in or more has been necessary for cool electric demand through the use of cool storage to be feasible. storage can mean significant savings. · The design team lacks experience · Backup or redundant cooling ca- or funding to conduct a thorough de- pacity is desirable. Cool storage can sign process. The design team should provide short-term backup or reserve be capable of TES design, which differs cooling capacity for computer rooms from standard HVAC system design. If and other critical applications. this is not the case, or if funding for de- sign fees is limited, the chances for a · Cold air distribution can be used, successful system are reduced. is necessary, or would be beneficial. Cool storage technologies using ice Performance testing of a new thermal permit economical use of lower- storage plant is particularly important. temperature supply water and air. Engi- Each system should be tested for neers can downsize pumps, piping, air charge capacity, discharge capacity, handlers, and ductwork, and realize and scheduling and control sequences. substantial reductions in first cost. The charge capacity test verifies that If one or more of the following are true, the system can fully charge storage TES may not be an appropriate tech- within the available time. The discharge nology: capacity test verifies its ability to pro- vide the required cooling, at or below

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 8 the maximum usable supply tempera- tentially at higher efficiency. And opera- ture, for each hour over the design load tion is generally at night when lower profile. Of primary concern is that stored ambient temperatures make for cooler thermal energy is used at a rate con- condenser temperatures, reducing en- sistent with the design. Especially in ergy use. chilled water systems, if oversized pumps circulate water at higher than However, TES systems do not always design flow rates, a smaller supply- save energy. A retrofit using ice storage return temperature difference will re- that fails to take advantage of the colder duce the capacity of the storage system. water available from the ice system may consume more energy than a direct A test of scheduling and control se- cooling system. Yet this system can quences confirms the proper operation make economic sense—consuming of valves, resetting of setpoints, and more off-peak, lower cost energy at starting and stopping of equipment, ac- night can still significantly reduce ex- cording to scheduled operating modes. pensive electrical demand and expen- sive on-peak electricity use. A TES system operates differently from a nonstorage system, in that the inven- Related to Implementation Cost tory of stored cooling must be properly managed. Operators of TES systems Most TES systems cost more up front must receive training in basic concepts and (if cost-effective) pay off through as well as the intended operating se- reduced electricity bills. But carefully quences and specific operating proce- designed systems needn’t cost more— dures for their systems. As operators or much more—than conventional gain experience they can improve sys- HVAC systems. tem performance and minimize operat- ing costs by refining the design operat- Sometimes cool storage systems can ing strategies and control setpoints. cost less and save energy. The Texas Instruments’ case study discussed ear- Related to Energy Savings lier describes one of three chilled-water storage installations in Dallas that have In new construction, TES systems can reduced energy consumption by more reduce overall energy consumption than 10 percent compared to nonstor- even though there may be increased age alternatives. The TI installation also use in the chiller: distribution of colder reduced capital costs millions of dollars chilled water can allow use of smaller by deferring the need for added chiller pumps and less fan horsepower to cir- capacity. Other factors that may help culate a smaller quantity of air through reduce cost of implementation include: the cooling coils. · Smaller air flows will provide ade- In retrofits it also is possible to reduce quate cooling in new construction or energy consumption but generally less major retrofits, especially for ice stor- so than in new construction. Here en- age. This can decrease the duct di- ergy savings result from chillers oper- ameter, reducing not only the cost of the ating at full load nearly all the time, po-

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 9 ductwork, fans, and pumps but also the Standard Savings Calculation installation labor cost. The following equation can be used in · Smaller ducts may increase rent- estimating demand savings from TES able space in new construction. systems. While energy consumption may increase or decrease, the change · System sizing is critical. For ex- generally has only a minor impact on ample, a careful assessment of the project economics. The time period number of hours that peak load must be when the energy is consumed has a met with stored cooling could show that significant impact; i.e. on-peak demand a considerably smaller storage system reduction or displacing expensive on- may only minimally affect demand re- peak energy with less expensive off- duction. peak energy use. In general, system economics depend heavily on the de- Estimation of Energy Savings mand savings and/or on-peak/off-peak rate differential. Demand savings must Demand costs are the main considera- be calculated for each month of chiller tion in determining the economics of a operation. TES system. Energy savings may be achieved, particularly in new construc- kWsavings = # tonsshifted tion or major renovation projects, but ´ (kW/ton)chiller performance typically are a small percentage of op- erating cost savings. In fact, energy Energy shifted to off-peak is more diffi- consumption may increase and still al- cult to calculate without monitored field low for an economically viable project. data or a calibrated, hourly computer simulation. A rough estimate can be To determine economic feasibility, accu- calculated as follows: rate cooling load data from the existing chiller plant is always best. If data are kWhshifted = # tonsshifted not available, not considered reliable or ´ (kW/ton)chiller performance incomplete, or if only a short time period ´ on-peak hours of data has been collected, an hourly ´ load shape factor computer simulation model must be de- veloped by a reputable energy analysis The load shape factor is a needed mul- professional. Available data, even for tiplier because peak cooling load typi- only a short period, can be used to help cally is not constant. This factor, used in calibrate the simulated building model the above equation, is for the on-peak and improve its accuracy. period only (the time when cooling load will be shifted) and for the peak cooling Back-of-the-envelope calculations can load for that day. Typical load shape give a rough estimate of possible sav- factors are in the range of 60 to 90 per- ings but must not be used for the final cent for a variety of building types and economic analysis, or chiller or storage climates. Annual energy shifted is the tank sizing. sum of daily energy shifted. At this point, an estimate can be made using

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 10 an average cooling load for each month a specific application can be obtained and the number of cooling days in the from contractors or vendors. month, then summing the monthly to- tals. Costs for chilled water tanks are based on volume, so the cost per ton-hour de- pends on the chilled water temperature Cost and Service Life range. Unit storage costs decrease as tank size increases. The cost of chillers or refrigeration equipment must be con- Factors That Influence Service sidered along with the cost of storage Life and First Cost capacity. Chilled water and phase- change material storage are compatible with typical conventional HVAC tem- Typically, a TES system increases costs peratures, and can often be added to compared to those for a direct cooling existing systems with no chiller modifi- system. But a much larger picture needs cations. For ice-harvesting systems, low to be looked at. Additional issues in- storage cost is offset by a relatively high clude: cost for the ice-making equipment.

· Floor Height: In new construction, Typical Service Life can low-temperature air distribution— which uses smaller, less expensive In addition to all equipment in a tradi- ductwork—reduce floor-to-floor height? tional cooling system, a TES system has a storage tank, pumps, piping and · Electrical Capacity: Can using possibly an interface heat exchanger. TES reduce the capacity and therefore The service life of each component the cost of the electrical service to a (except the actual storage tank) has new project, or avoid increasing the been estimated and can be found in service in the case of a building expan- Reference 1. All TES-specific compo- sion? nents are rated at a 20-year minimum service life. Storage tanks, generally · Useable Space: Can using TES, concrete or steel, also have service with an underground storage tank, such lives of at least 20 years. as under a parking lot, free up space in an existing chiller plant or reduce the Operation and Maintenance size of a new structure? Requirements

Costs given in Table 1 are general Factors that tend to increase mainte- guidelines for initial economic evalua- nance costs for cool storage systems tions of storage systems. They include compared to nonstorage systems in- the cost of storage tanks and any re- clude: quired internal diffusers, headers, or heat transfer surface. Costs will vary · Annual tests to ensure solutions depending on the size of the project and contain proper coolant concentration, site-specific considerations, among levels of corrosion inhibitors, and other other things. Accurate cost estimates for

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 11 Chilled Ice External Internal Encapsulated Phase-change Water Harvester Melt Ice Melt Ice Ice Material Pre-packaged or Low-temperature Low-temperature Low-temperature built-up ice- Chiller Cost Standard water coolant or built-up secondary secondary Standard water making refrigeration plant coolant coolant equipment Chiller Costa $/ton 200-300, 1,100-1,500 per 200-300, 200-500 200-500 200-500 or use existing ice-making ton or use existing

$/kW 57-85 313-427 57-142 57-142 57-142 57-85 Tank Volume ft³/ton-hr 11-21 3.0-3.3 2.8 2.4-2.8 2.4-2.8 6.0 Storage Installed Costb $/ton-hr 30-100 20-30 50-70 50-70 50-70 100-150 $/kWh 8.50-28 5.70-8.50 14-20 14-20 14-20 48-43 Charging Temperature (oF) 39-42 15-24 15-25 22-26 22-26 40-42 Chiller Charging Efficiency kW/ton 0.60-0.70 0.95-1.3 0.85-1.4 0.85-1.2 0.85-1.2 0.60-0.70 COP 5.9-5.0 3.7-2.7 4.1-2.5 4.1-2.9 4.1-2.9 5.9-5.0 Discharge 1-4 above Temperaturec (oF) charging 34-36 34-36 34-38 34-38 48-50 temperature Secondary Secondary Discharge Fluid Water Water Water Water coolant coolant Open or closed Tank Interface Open tank Open tank Open tank Closed system Open tank system Use existing High High Modular tanks Tank shape Use existing Strengths chillers; fire instantaneous instantaneous good for small or flexible chillers protection duty discharge rates discharge rates large installations Separate charge Storage capacity Requires and discharge increases with clearance above Comments circuits. Charge larger temperature tank for ice with coolant or range. maker. liquid refrigerant. Notes: a: Costs are for chiller or refrigeration plant only, and do not include installation. All costs, except ice harvesters, are per nominal ton. Derating for actual operating conditions may be required. b: Costs are for storage only, and include tank, internal diffusers, headers, and heat transfer surface. c: Typical minimum temperatures, with appropriate sizing of storage capacity. Higher temperature can be obtained from each medium.

Table 1: Comparative Costs and Performance of Cool Storage Systems (Source: ASHRAE )* additives for ice storage systems using · Added maintenance for cool stor- glycol or other secondary coolants. age system components, such as addi- Some glycol manufacturers provide free tional pumps, heat exchangers, and laboratory analysis of samples. There control valves. may also be an expense associated with replacing glycol lost during mainte- Factors that tend to decrease mainte- nance procedures and through leaks. nance costs for cool storage systems include: · Increased water treatment ex- pense in chilled water storage and · Smaller components, such as chill- some ice storage systems, which con- ers, pumps, and cooling towers, for typi- tain large volumes of chilled water in the cal systems. tanks.

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 12 · Glycol or other secondary cool- · Demand Charge: A tariff added to ants in the system provide coil freeze a customer’s electric bill that increases protection, and eliminate the need to in proportion to maximum kilowatts drain the cooling system in the winter, used. or to use special controls to prevent coil freeze-ups. · DX (direct expansion): Refers to a heat exchanger that contains the refrig- · Cooling loads can be met from erant inside its tubing rather than water, storage while some equipment is taken antifreeze, or other fluid. Heat from the out of service for maintenance in some surroundings is directly absorbed into storage systems. the refrigerant, which is “pumped” by the compressor.

Laws, Codes, and · Figure of Merit (FOM): A measure Regulations of a storage tank’s ability to maintain separation between warm and cool wa- ter. All equipment and components used for a TES system should conform with the · Full Storage: Refers to a TES sys- same laws, codes, and regulations re- tem that stores sufficient cooling to meet quired for traditional cooling systems. an entire peak day cooling capacity, al- Large tanks may pose a problem. Zon- lowing chillers to be off during the on- ing requirements, particularly height re- peak period. strictions, should be checked early. If height is an issue, it is possible to com- · Off-Peak: A time period, defined by pletely or partially bury it. It is also pos- the utility, when the cost of providing sible that a levee will be required power is relatively low, because the around the tank in case of a rupture. system demand for power is low. The off-peak period is often characterized by If a fire-protection storage tank is re- lower costs to the customer for energy quired on site, as it would be at some costs, and either no or low demand manufacturing facilities, it may be pos- charges. sible to use this tank to store chilled water for the cooling system. If this is · On-Peak: A time period, defined by available, it should be carefully checked the utility, when the cost of providing to ensure code compliance. power is high because the system de- mand for power is high. The on-peak period is typically characterized by Definitions of Key Terms higher costs to the customer for energy and/or demand charges. · Condenser: The container in a cooling system where gas changes · Partial Storage: Refers to a TES phase to liquid, releasing heat to the system that contains sufficient cool surroundings. storage to meet part of the cooling load of a facility. Such systems are used in

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 13 conjunction with a chiller system to pro- 4. E Source, “State of the Art Technol- vide required cooling capacity. ogy Atlas: Commercial Space Cool- ing and Air Handling,” Chapter 11, · Phase Change: As a substance 1995. changes between its solid, liquid, and gaseous forms, it is said to change 5. International Thermal Storage Advi- phase. During transitions in phase— sory Council (ITSAC), 3769 Eagle freezing, melting, condensing, boiling— Street, San Diego, CA 92103, Tel: the material releases or absorbs large (619) 295-6267. amounts of thermal energy without changing temperature. The energy as- sociated with this is called latent heat. A Major Manufacturers material that can store thermal energy as latent heat is called a phase-change Chicago Bridge & Iron Co. material. 1501 N. Division Street Plainfield, IL 60544-8929 · Sensible Heat: Heat that can be Tel (815) 439-6000 perceived by the human senses, as op- Fax (815) 439-6010 posed to latent heat (see phase change E-mail: www.chicago-bridge.com definition). San Luis Tank & Piping Co. 825 26th Street References to More Paso Robles, CA 93447 Information Tel (805) 238-0888 Fax (805) 238-5123

1. American Society of Heating Refrig- Cryogel eration, and Air-Conditioning Engi- P.O. Box 910525 neers, “ASHRAE Handbook HVAC San Diego, CA 92191 Applications,” June 1995. Tel (619) 792-9003 Fax (619) 792-2743 2. American Society of Heating Refrig- E-mail: [email protected] eration, and Air-Conditioning Engi- neers, “Design Guide for Cool For more information on companies who Thermal Storage,” 1994. make TES components see Reference 4 above. In addition, you may contact 3. Electric Power Research Institute the International Thermal Storage Advi- (EPRI) HVAC & R Center (formerly sory Council (Reference 5) or the EPRI the Thermal Storage Air Condition- HVAC & R Center (Reference 3). ing Center), University of Wisconsin, 150 East Gilman Street, Suite 2200, Madison, WI 537093, Tel: (800) 858- 3774, Fax: (608) 262-6209. *Reprinted with permission. Copyright, 1994 by ASHRAE. All rights reserved.

PG&E Energy Efficiency Information© “Thermal Energy Storage” Page 14