District cooling and thermal energy storage – new markets emerge

Although most existing installations are in North America, new markets are opening up around the world for district cooling plants. John S. Andrepont explains how, for high density load situations, district cooling can easily out-perform conventional cooling technologies – particularly when used alongside a thermal storage system to reduce the impact of peak loads.

Two complementary technologies – district cooling and thermal energy storage – are being increasingly used, not only in their traditional US market but also in other markets around the world.


District cooling (DC) involves the provision of cooling to multiple buildings or facilities from one or more central cooling plants that are interconnected to the cooling users via a network of supply and return piping. Most DC systems distribute chilled water, though some distribute other chilled fluids, or occasionally even refrigerant.

The Middle East has strong potential for incorporating district cooling in new building developments
Click here to enlarge image

The central plants in DC systems can utilize electric driven chillers, heat-driven chillers (e.g. absorption or steam turbine-driven chillers), or a hybrid combination of electric and non-electric chillers. The business model of a DC system commonly fits into one of two forms:

  • Single owner-user DC systems. In this case, a single business entity owns and operates the DC plant and distribution piping system, as well as the facilities being cooled by the DC system. Typical examples include college and university campuses, medical complexes, airports, and military and other (federal, state and local) government building complexes, as well as some large private commercial campuses and industrial facilities.
  • DC thermal utility systems. In this case, one business entity (the DC system developer-owner) owns the central cooling plants (and typically the distribution piping network), while one or more cooling users (up to dozens or even hundreds) are independent business entities. In general, the cooling is provided by the DC system owner to the independent users under long-term (often 20 years or longer) service contracts. Typical examples are urban DC systems, planned or redeveloped mixed-use communities, and/or outsourced cooling services for any of the types of systems otherwise served by a single owner-user DC system.

Both families of DC business models use similar technologies, though the commercial details can be quite different.

Drivers for the increased utilization of DC generally include one or more of the following:

  • increased demand for air-conditioning
  • large-scale real-estate developments
  • improved life-cycle economics
  • enhanced reliability of the cooling service.

The second of two chilled water TES tanks at the district cooling system serving St Paul, Minnesota. North America currently has the largest existing district cooling market (Chicago Bridge & Iron Company)
Click here to enlarge image

DC systems can be purely cooling systems, or they can be configured and operated in concert with district heating (DH) systems. DC or district heating and cooling (DH&C) systems can also be integrated with on-site power generation, cogeneration or trigeneration to provide a CCHP (combined cooling heating and power) system. Finally, DC systems are often configured to utilize cool thermal energy storage (TES).


Thermal energy storage (TES) has a long, successful history of use in large air-conditioning systems including private commercial/industrial, public/institutional and utility DC applications. TES can economically and efficiently shift peak air-conditioning and other cooling loads from on-peak to off-peak periods.

TES technologies for cool storage include two distinct types:

  • latent heat storage systems such as ice TES in which thermal energy is stored as a change of phase of the storage medium (usually between solid and liquid states)
  • sensible heat storage systems such as chilled water (CHW) TES and low temperature fluid (LTF) TES in which thermal energy is stored as a temperature change in the storage medium used.

Each TES technology has inherent advantages and limitations which affect its performance and suitability in certain situations. An understanding of these characteristics is critical in selecting and optimizing the type and configuration of TES for any particular application. Table 1 presents a summary of the characteristics of the various TES technologies; it is intended merely as a starting point in the proper application of TES.

Click here to enlarge image



The findings of two recent surveys of TES installations – one for TES use in campus DC applications1 and one for TES use in DC utility applications2 – illustrate the extensive and growing application of TES. The databases developed in these surveys also document the demographics of TES use in the campus DC and DC utility areas, noting specifically:

  • numbers and capacities of TES installations
  • geographical distribution
  • chronological distribution and growth trends
  • TES technology types
  • frequency of instances of phased capacity additions of TES
  • owners with TES installations in multiple DC systems.

The surveys identified most of the US applications but only a smaller sampling of non-US applications, and are likely to have significantly understated the latter. Highlights from the surveys are presented in Table 2, while Table 3 shows the geographical distribution of DC utilities in countries other than the US.

Click here to enlarge image


Click here to enlarge image

In addition to the instances of DC systems with TES identified in these two recent studies, there has been rapidly increasing DC activity – especially outside the US and mostly DC with TES. In just the past two years, numerous new (and expanding) DC systems, most with new TES elements, have been established or planned in many countries around the world (Table 4), including many with multiple developments. The most active region for growth has been the Middle East.

Click here to enlarge image

Various factors are behind the increased use of DC. Although DC is quite capital-intensive (because of the cost, not only of the central cooling plants, but also of the distribution piping networks), it generally has a life-cycle cost benefit compared with individual building cooling systems. Where decision-makers focus on energy efficiency and life-cycle economics, DC will often have an advantage.

This is true for both single owner-user DC systems and developers of DC utility systems. The customers of DC utility systems actually gain not only the life-cycle cost benefit, but also avoid capital costs as these are absorbed by the DC utility in exchange for the long-term customer payments for the cooling service.

In the most active areas (e.g. East Asia and especially the Middle East), DC growth is being driven by local economic expansion and the related investments in major (and mega) real estate developments (e.g. the many artificial island communities being built in Dubai). These large-scale developments are generally planned centrally, with much forethought being given to the supporting infrastructure. Planners and developers find that DC is a natural adjunct to other traditional centrally planned utilities such as electricity, gas, water, sewerage, telecom, etc.

Reliability of cooling is another of DC’s strong points. Unlike individual building cooling systems, where it is not often practical to incorporate redundancy, DC systems routinely have installed back-up capacity coupled with high reliability and quality of service.

DC is also growing in areas of Europe where DH has been routinely practised for many decades but where the population now expects to have summer air-conditioning when it was once a luxury.

In many instances, DC is the technology of choice because of its ability to provide energy efficiency and environmental benefits. This is especially true when DC is complemented by other technologies such as hybrid chiller plants, TES, DH, combined heat and power (CHP) or CCHP, and even renewable ‘free’ cooling from deep cold water sources – all technologies generally not practical or economical on an individual building scale.


Traditionally, the main benefit associated with the use of TES has been reduced peak power demand and, therefore, reduced operating energy costs resulting from lower demand charges and/or lower time-of-use energy charges. However, there are many other, sometimes even more significant, benefits of TES. These other benefits are key drivers in the increased use of TES in many DC developments, worldwide.

The value of load shaping

Even with ‘flat rates’ for energy charges, there is real value in the use of TES by energy consumers. The reality is that the cost and value of electrical energy varies continuously (based on supply and demand). The variability between day and night is significant, with the incremental power plant costs during peak demand periods in the daytime being far more than the costs during low demand periods at night. The providers of energy continue to realize these variable costs even if they charge their customers a ‘flat rate’ for energy.

Using TES to ‘shape’ the load usage profile – specifically to minimize energy consumed during high demand (high value) periods and to maximize use during lower demand (low value) periods – makes the energy consumer a more attractive buyer to the energy provider. The energy provider can offer a substantially lower ‘flat rate’ to those consumers with these more attractive load shapes as it costs less to serve them.

Net capital cost reductions with TES

In large-scale applications, such as DC systems, the economy-of-scale of TES (especially sensible heat TES systems using CHW or LTF storage) often produces immediate net capital cost savings compared with the equivalent conventional (non-TES) chilled water capacity. In particular, the net capital cost saving can be captured from using TES at times of:

  • new system construction
  • retrofit expansions
  • chiller plant rehabilitation.

In each case, the savings occur because TES allows the installed chiller plant capacity to be ‘downsized’; instead of needing chillers equal to the peak day’s peak load (plus any necessary spare capacity), the chillers need merely be equal to the peak day’s average load (plus any necessary spare capacity). Additionally, in large applications, the cost of TES is less than the avoided cost from downsizing the chiller plant. Numerous examples of millions of dollars of net capital cost savings have been documented3 (some of these are summarized in Table 5).

Click here to enlarge image


Efficiency and emission benefits of TES

Of course there are certain inefficiencies inherent in TES – specifically the added pumping energy associated with charging and discharging TES and heat gain into TES. But, in most applications, these inefficiencies are relatively small and largely (or more than) compensated for by efficiency improvements. These improvements are associated with:

  • operating chillers more at night, when the more favorable heat rejection conditions reduce chiller plant power consumption
  • avoiding having to operate chillers and their auxiliaries at severely low part loads – a condition which normally results in high unit power consumption for many hours of the year
  • in systems incorporating ‘free cooling’, increased hours per year to utilize that ‘free cooling’.

In a number of chilled water TES applications, net savings of 3%-25% in on-site annual kWh per ton-hour for chilled water production have been documented, with the reduction in most cases being 8%-13%.4 Even more significantly, a TES system employed at an electric energy consumer’s facility typically produces significant savings in energy (fuel consumption) and emissions at the source power plants.

This is related to the fact that a power plant ‘on the margin’ during the daytime hours of peak power demand has a much higher ‘heat rate’ (i.e. a much lower fuel efficiency) than one ‘on the margin’ during the night-time hours of lower power demand.

Independent studies carried out in California for two utilities,5 in Texas for one utility6 and in Wisconsin for two utilities7 identified dramatic reductions in energy and emissions by shifting energy use from off-peak to on-peak periods. Annual source energy (fuel) use was typically reduced by 5%-24% (most often 15%-18%), while annual emission of air pollutants (sulphur dioxide and nitrogen oxides) was reduced by 5%-29% (most often 16%-22%) and annual emission of carbon dioxide (a greenhouse gas) was reduced by 5%-24% (most often 11%-17%).

Economical energy storage

Most systems, whether natural or man-made, use some form of storage to beneficial effect. The electric power industry (suppliers and users alike) would enjoy enormous benefits from an ability to store energy, but there are very limited options for the economical storage of electricity.

Pumped storage hydroelectric plants are effective in very large-scale applications, but they typically exhibit parasitic energy losses of 30% or more during charging/discharging, cost thousands of dollars per kW, and take many years to permit and build (if they can be sited and permitted at all). Compressed air energy storage (CAES) is also very high in terms of capital cost, has high parasitic losses, limited siting options, and is still somewhat developmental. Flywheel storage is also high in cost and quite developmental. Battery technology is costly, somewhat developmental, and as yet deployed only in relatively small scale; yet it is often considered the most economical and near-term means of commercially ‘storing electrons’.

However, TES can be employed in both demand-side and supply-side applications of DC, is already fully commercial, and deployable at quite reasonable cost.8


TES effectively decouples the generation of cooling from the demand for cooling. Thus TES adds an entirely new degree of freedom (flexibility) to the manner in which cooling systems can be operated.

Some TES system designs provide more flexibility than others. When facing the challenges of evolving energy markets, the systems with the greatest flexibility will provide the greatest ability to capture and maximize value. The examples below serve to illustrate some of the various types of flexibility that can be provided by TES.

  • TES systems that can be rapidly discharged (rather than merely discharged over a prescribed period of a traditional time-of-day electric utility tariff) can be controlled to vastly increase benefits in certain operating scenarios.
  • TES systems designed for a ‘full shift’ of current system loads can operate as a ‘partial shift’ of larger future loads. Conversely, systems designed for a ‘partial shift’ or ‘load levelling’ of peak day loads can operate as ‘full shift’ systems on lower load days, or perhaps as ‘full shift’ systems on peak days, albeit for a shorter period of time (if warranted by electricity pricing).
  • Many TES systems are now routinely pre-designed for possible conversion from initial CHW TES service at conventional operating temperatures to LTF TES service at lower supply temperatures and higher temperature increments ( T) to dramatically increase TES capacity to meet either larger future loads and/or more volatile real-time energy price signals. Capacity (storage and discharge rate) increases of 40%-70% are common.
  • The campus DC system at Princeton University uses a 40,000 ton-hour LTF TES system in conjunction with a hybrid chiller plant, a CHP system and real-time energy purchases. By fully discharging TES in as little as four hours and sometimes discharging up to 1.5 tank charges per day, TES plays a key role in maximizing energy cost savings in a volatile market. The available low fluid supply temperature from TES (0à‚ºC) also provides flexibility in selectively lowering the chilled water supply temperature (and thus raising the deliverable cooling capacity) in the campus DC network.
  • The recently expanded DC system serving DFW (Dallas/Fort Worth) International Airport incorporates a 90,000 ton-hour LTF TES tank. Designed to fully discharge in as little as three hours, it generates more savings than originally anticipated. Although its energy is purchased on a ‘flat rate’ basis, DFW does incur a demand charge related to its transmission and distribution (T&D) billing. But this demand charge is not based on DFW’s peak power demand each month, but on the power demand that occurs coincidently with the monthly peak demand of the local electric grid. Because the local grid exhibits a monthly peak that occurs very predictably and repeatedly within the same narrow time window in the late afternoon, DFW has compressed the time during which it discharges TES to coincide with that time window. In doing so, DFW doubled its demand savings, nearly achieving a full load shift, and realizing a reduction in demand charges of 85%.
  • Many TES installations can serve a dual-use; for example, CHW TES can also act as a fire protection water reservoir, reducing risk and (often) reducing insurance premiums.
  • TES of all types can serve as redundancy and emergency back-up cooling capability (increasingly required in data centres or other ‘mission critical facility’ designs).
  • CHW TES can often extend the annual hours of usage from ‘free cooling’ cooling tower systems.
  • CHW TES can economically extend the peak capacity of deep water source cooling systems.


As described in the May-June 2006 issue of Cogeneration & On-Site Power Production, there are often advantages for CHP and on-site power systems when those systems are coupled with DC and TES. DC provides a sizeable local thermal load that can be served by a CHP or other on-site power system. These cooling loads can be served from an on-site power system either directly by heat-driven chillers (e.g. steam turbine-driven compression chillers or absorption chillers using steam, hot water or direct exhaust heat) and/or indirectly by electric-driven chillers.

In addition, complementing the DC system with TES means that the cooling demand versus time-of-day relationship can be shaped to better match the coincident electric and thermal outputs of a CHP system, thus improving its economics.

Finally, there is another synergy with DC, applicable where the on-site power system uses one or more gas turbines (which have the inherent characteristic of losing power output as inlet air temperature increases). In such situations, the cooling system can be used to provide cooling of the gas turbine inlet air, typically providing an economical increase in hot weather power output of 20%-25% or more.


The continuing expansion of DC systems, and DC with TES, can be confidently predicted in worldwide markets wherever growth in air-conditioning loads is occurring. For relatively high- density collections of cooling loads, life-cycle economics favour the use of DC versus individual building chiller systems. Coupling TES with DC also often improves both the operating and capital costs. Expect to see more DC and DC with TES.

John S. Andrepont is the founder and President of The Cool Solutions Company, providing consulting services in the areas of thermal energy storage, district cooling and turbine inlet cooling, based in Illinois, US.
e-mail: CoolSolutionsCo@aol.com


1. Andrepont, J.S., Cool trends on campus: a survey of thermal energy storage use. District Energy, 91(1), 25-30. First Quarter 2005.

2. Andrepont, J.S., Thermal energy storage (TES) in district cooling utilities – overview and findings from a survey of applications – a presentation preceding a panel discussion at the IDEA Annual Conference, June 2005.

3. Andrepont, J.S., Reducing energy costs and minimizing capital requirements: case studies of thermal energy storage (TES). In: Proceedings of the IETC (Industrial Energy Technology Conference), May 2007.

4. Andrepont, J.S., Thermal energy storage: analysis and optimization of campus cooling systems. In: Proceedings of CAPPA (Central APPA) Technology 2001, February 2001.

5. Tabors Caramanis & Associates, Source energy and environmental impacts of thermal energy storage. Report for the Thermal Energy Storage Systems Collaborative of the California Energy Commission, September 1995.

6. Reindl, D.T., et al., Characterizing the marginal basis source energy emissions associated with comfort cooling systems. Report No. TSARC 94-1, University of Wisconsin HVAC&R Center, Madison, Wisconsin, December 1994.

7. Gansler, R., Energy and environmental impacts of space conditioning systems. Report on ASHRAE Research Project 991-RP, University of Wisconsin HVAC&R Center, Madison, Wisconsin, August 1999.

8. Andrepont, J., Energy storage: not just R&D nor necessarily expensive. Power Engineering, 110(5), 56. May 2006.

No posts to display