Figure 1. Full storage system.

Some thermal storage media are familiar to most of us, but what about gravel? Or lake water? See how those can work in the right environments, and tuck this away as a general refresher on thermal storage concepts and options for use in hospitals, data centers, and elsewhere.

Thermal storage systems include any process that can capture energy at a point in time, store the energy, and then release the energy at a later point in time to heat or cool. Thermal production can be a natural process, a mechanical process, or a chemical process. Thermal storage provides either heating or cooling for a building, a group of buildings, manufacturing, or other processes. 


There are many benefits to using thermal storage. Here, we will provide an overview of some of the many benefits of thermal storage. Other benefits include dual use as fire protection as well as benefits to utility companies and society as a whole.


Facilities can reduce their on-peak (and even mid-peak) power demand (typically the most expensive utility rate period) by using thermal storage. As shown in Figure 1, full thermal storage systems are sized to provide all the cooling required during a certain part of the day (generally on-peak hours) with the cooling equipment shut off. During non-peak hours, cooling equipment operates at a higher capacity than is required to meet the instantaneous cooling load, generating cooling surplus to be stored.

Figure 2. Partial storage system.

As shown in Figure 2, partial storage systems are sized so that on peak-cooling days, the smallest chiller plant possible can meet the peak-cooling load by operating at a constant rate throughout the day in conjunction with the thermal storage system. That is, partial storage requires the chiller plant to operate during the entire day, including during the on-peak period.


A good storage system can provide better energy performance than a conventional instantaneous system because of the high Delta-T and its beneficial effect on series chillers, and because lower wetbulb temperatures at subsequent lower condensing temperatures are available at night. Increased efficiency also results from all the chillers (and their auxiliaries) operating closer to design points more of the time. Inefficient, low part-load chiller operation is unnecessary.

Figure 3. Multiple tank system.


Lower facility electric demand results in lower utility demand costs, which are often a significant portion of the lower energy costs experienced by facilities with installed thermal storage systems. However, reducing energy consumption and shifting production from expensive utility energy rate periods to less expensive utility energy rate periods also lower facility energy costs. Based on experience, annual energy cost savings can be as much as 40% depending upon the thermal storage system installed. Installations with real-time priced power have shown even greater savings.


Maintenance costs are lower because thermal storage systems typically use smaller chillers, cooling towers, and pumps than do conventional systems. While increased hours of operation may offset these savings to some extent, stops and starts affect maintenance more than hours of operation. Also, typically, a conventional system runs most of the year in partial-load conditions, which is just as demanding on maintenance as full-load operation.


A properly planned and designed thermal storage system often provides lower capital costs in addition to lower energy use and lower energy costs. This is due to smaller equipment requirements and smaller related infrastructure requirements, which is possible even with the added cost of the thermal storage tank. For example, a thermal storage system may only require 1,000 tons of cooling equipment vs. 2,000 to 3,000 tons of cooling equipment required for a conventional cooling plant with no thermal storage. Not only does the cooling production equipment cost much less, but the electrical and building infrastructure costs to support the cooling equipment will also be lower, potentially more than offsetting the capital cost of the thermal storage system, thus producing a net capital cost savings vs. a non-storage system.


Thermal storage can play an important role during an emergency. Some facilities have critical cooling (or even heating) loads, where even a small outage could be devastating. For example, a data center containing key financial or medical information requires continuous cooling and could not tolerate a cooling outage even for 10 to 20 min while the cooling plant was brought back online.

With some modest emergency power generation for chilled water pumping, thermal storage could be designed to provide 20 min of peak cooling to a data center, allowing the chiller plant to be brought back online during an outage. Likewise, a thermal storage system could provide emergency cooling to surgery rooms and specialized equipment for a period of time during an outage while cooling equipment was made operational and/or while operations/systems are safely shut down.


There are many different types of thermal storage systems, which are divided into two broad groups: sensible thermal storage systems and latent thermal storage systems. Sensible thermal storage systems achieve thermal energy storage (TES) by a difference in a material’s or substance’s temperature, while latent thermal storage systems achieve TES by a change in a material’s or substance’s phase. Sensible systems are subdivided into solid thermal storage systems and liquid thermal storage systems.


Solid sensible thermal storage systems use a change in a material’s temperature to store thermal energy (or technically, in the case of cooling TES, to store an energy sink). There are three typical types of solid thermal storage systems: use of building mass, use of the ground, and use of a specially constructed system.


Building mass is commonly used for thermal storage without even considering that thermal storage is being done or why the process works. Some buildings are ventilated with outside air at night to cool the building mass. The colder building materials are able to absorb some of the daytime heat gains, reducing the amount of mechanical cooling needed during the day.

Figure 4. Background stratified chilled water tank section.

Many buildings with large fluctuations in load use the building mass to help meet peak cooling loads. Large restaurants, meeting halls, and even sports arenas often “pre-cool” the space to reduce the peak cooling load. The spaces start out somewhat cooler than optimum, and often finish somewhat warmer than optimum; but much less cooling capacity is needed, and electrical demand is reduced.


The ground has been used for thermal storage, especially for longer-term (such as annual) thermal storage. The goal of such thermal storage systems is usually to heat the ground with “free” or recovered heat, including space heat rejected during the warmer part of the year. The stored heat is recovered and used during the colder part of the year. This concept can reduce the winter heat needed and at the same time reduce the summer mechanical cooling needed.

Thermal storage systems have also been designed to cool a transfer medium, such as water, with cold winter air and circulate it in the ground to cool the ground. The systems then reverse in the summer to draw on the stored cooling from the ground, and the heat from the building is stored in the ground. Most often this concept requires two or more wells and reversing operation, depending on whether the system is being used for heating or cooling. To be successful, a ground mass thermal storage system cannot be a great deal warmer or colder than the normal earth temperature, since the greater the temperature differential, the faster/greater the associated heat loss.


Gravel, either in a constructed enclosure or in the ground as a functional structural material, is used for thermal storage. Many systems have been designed to use this type of thermal storage. Most of them are either a part of a solar thermal system, or a “free” cooling system. With gravel thermal storage systems, air heated by the sun is forced by a fan or by natural convection through a large bed of gravel, which heats the gravel during the sunlight period. At nights or during cloudy weather, the airflow is reversed and heat is delivered to the building. Gravel is also used to store cooling. Cold night air is drawn through the gravel storage to cool the material. Airflow is reversed, when cooling is needed by the building served.


As with solid sensible thermal storage systems, liquid sensible thermal storage systems use a change in a material’s temperature to store thermal energy. Liquids are more commonly used for sensible thermal storage than solids. One important design consideration of a liquid storage system is the need to maintain a separation between the colder fluid and the warmer fluid. There are several ways the challenge of temperature separation has been addressed including: natural storage systems, multiple tank systems, and stratified thermal storage tanks.


Nature provides the container for some liquid thermal storage systems. One such system, where winter weather is very cold, uses a lake for thermal storage. Nature cools the lake water, and the cold surface causes the water to reach its maximum density point (about 39°F) and sink to the bottom of the lake.

In the summer, when cooling is required by a nearby university campus, cold lake water circulates from the bottom of the lake through plate-and-frame heat exchangers and returns to the warmer top of the lake. The large mass and depth of the lake is essential to providing annual thermal storage in this way, as the surface of the lake naturally warms during the spring, summer, and fall period, while the deep lake water remains cold year round. This lake cooling system is able to provide campus cooling with pumping energy only, which is approximately one order of magnitude less than required for mechanical cooling.


As shown in Figure 3, this type of system uses multiple tanks for storage, including one more tank than the actual designed liquid storage volume. For example, a system with 10 tanks holding 10,000 gal each requires an extra 10,000-gal tank. Valves on each tank direct which tank the liquid will be drawn out of and which tank will serve for the return water. As one tank is being filled, another tank is being emptied. Once a tank becomes empty, the valves are adjusted to draw water from a previously unused full tank and return the water to the new empty tank.

The multiple tank concept eliminates mixing of supply and return water, but it does require extra storage volume, higher unit capital cost (associated with smaller tanks), higher complexity of piping and valves, more pumping horsepower (other factors being equal), and it introduces air into the system as water drops into the top of an empty tank, which can increase corrosion and microbiological activity.


By far the most common sensible thermal storage systems today are stratified thermal storage systems. The popularity of vertical stratified thermal storage systems is due in part to the simplicity of the system. And while the basic stratified thermal storage system is simple in concept and construction, the application of the system to real life situations requires a great deal of care in planning and design. As a consequence, stratified thermal storage is most often applied to larger systems with individually engineered designs.

Stratified thermal storage depends on the density difference between colder and warmer water. This temperature stratification has been used for centuries in hot water storage, but more recently it has formed the preferred solution in chilled water storage. The big difference between hot water thermal storage and cold water thermal storage is the relative power of the density differential to naturally separate supply from return water.

In hot water or heating water systems, there is usually a very large temperature difference between the colder water and the warmer water. That large temperature difference results in a large difference in density, which provides substantial buoyancy for the warmer water to float on top of the colder water.

Figure 5. University Hospital TES tank.

Also, as water gets colder, it approaches its maximum density point (at about 39.4° or 4.1°C). The closer the water gets to the maximum density point, the less is the difference in density for each degree of temperature change. While this complicates things somewhat for stratified cooling water thermal storage, designers have found that, with careful tank and flow diffuser design, the tendency of chilled supply and return water to mix can be reduced. The simplicity of stratified storage designs has led to the concept being developed as the preferred solution.


A chilled water TES study was conducted for a university to determine the best method of meeting the growing demand for central plant cooling on the university campus. The university includes a growing educational health science institute, with academic programs featuring more than 100 degrees and certificate programs in the area of medicine, dentistry, nursing, public health, and allied health, as well as graduate studies. In addition to the teaching and research functions associated with the general campus, the university also maintains a hospital, children’s hospital, and emergency room.

Similar to many other institutions, the university continues to expand and therefore campus loads (cooling, heating, and electrical) continue to grow. When this study was conducted, peak electricity usage had surpassed the capacity of the existing central cogeneration plant, and the university was forced to purchase power from the utility company during most daytime periods, including the on-peak period. The electric demand resulting from the operation of the existing electric-drive chillers and ancillary equipment within the central plant was approximately 25% of the total campus demand. Implementing thermal storage provides a method of shifting cooling production from the on-peak period to the off- and mid-peak periods, reducing the amount of power that would need to be purchased and instead allowing the cogeneration plant to utilize excess electrical capacity during those periods.

Figure 6. TES tank under construction.

The existing peak cooling load was approximately 7,200 tons, and the total peak day cooling load was approximately 100,000 ton-hrs. The future peak cooling load was estimated to be approximately 12,500 tons and the total peak day cooling load was projected to be approximately 170,000 ton-hrs.

Three central chiller plant alternatives were generated to meet the existing and future needs of the university: a conventional chiller plant with no TES system, a full storage TES plant, and a partial storage TES plant. Energy modeling, budget construction cost estimates, lifecycle cost analysis, and net present value calculations were created for all three alternatives and compared to determine the appropriate solution for the university.

This study showed that partial storage was the most economical alternative, with an NPV of approximately $23.7 million, approximately $2.6 million of NPV savings over the conventional alternative. This was due in part to a smaller storage tank and smaller chillers, resulting in a smaller capital cost. It was determined that this alternative would provide sufficient cooling capacity to meet the projected cooling demands and significant energy cost savings for the university, and was, therefore, the recommended alternative. Construction was completed in 2011. Figures 5 and 6 show the TES tank.  ES