Industrial News
Current Position: Home > News > Industrial News
Ice Storage Retrofit for Rooftop Air Conditioning


A significant fraction of space in commercial and Federal buildings is cooled by single-package rooftop air conditioning units. These units operate during the day under hot conditions. Consequently, their energy efficiency, as compared with chiller systems for building cooling, is generally much lower. Several U.S. companies are developing rooftop units. Although the low evaporator temperatures needed for ice making tend to reduce the efficiency of the chiller, the overall efficiency of the ice storage system may be higher than the efficiency of a packaged, conventional rooftop installation. One version of this concept, the Roofberg® System developed by the Calmac Corporation, was evaluated on a small building in Oak Ridge, Tennessee.

Roofberg consists of a chiller, an ice storage tank and one or more rooftop units whose evaporator coils use a glycol solution for cooling. The ice storage tank contains coils of heat exchanger tubing through which a solution of ethylene glycol passes. Water fills the tank and surrounds the tubing so that ice is formed around the tubing. The ice storage process continues until most of the water in the tank has frozen. In the test system, about 16,000 pounds of ice are made and stored each night ready to meet the daytime cooling demand of the building. During the discharge process, ethylene glycol is circulated through the coils where it is cooled by the melting ice and to the rooftop units to provide cooling to the building. The ice storage component decouples the cooling demand of the building from the operation of the chiller. The chiller can operate at night (cooler, more efficient condensing temperatures) to meet a daytime cooling demand. This flexibility permits a smaller chiller to satisfy a larger peak cooling load. Further, the system can shift the cooling demand to off-peak hours when electricity from the utility is generated more efficiently and at lower cost.

Integrated Ice Storage/Rooftop Advantages

When considered as an alternative to rooftop replacement, there are some obvious advantages to integrating ice storage to existing rooftop systems. In multistory buildings or large single story buildings, it requires a crane or a helicopter to remove/replace an existing rooftop unit. Depending on the logistics, the rigging requirements for replacing rooftops can add appreciably to the cost. Second, it is likely that the original rooftop curbing will not be the right size for the new rooftop. Replacing old curbing with new will be an added expense which may be compounded by other costs. Third, ice storage permits the building owner to take advantage of time-of-use electric rates and demand cost reductions for smaller on-peak electric loads. And fourth, there can be savings in operating costs by substituting a chiller system that is more efficient than a replacement rooftop unit.

Ice Storage/Rooftop Field Study

The DOE Federal Energy Management Program/'s New Technology Demonstration Program sponsored a demonstration project of the Roofberg system. Oak Ridge National Laboratory (ORNL) staff conducted the project, which began in the summer of 1996 using an ORNL facility as the test site. The Calmac Corporation supplied one 190 ton-h ice storage tank (Roofberg®) and the Trane Company provided a 20-ton, air-cooled chiller. The building selected for the demonstration was cooled by five packaged rooftop units.

The ice storage system to be tested was installed as a retrofit to three of the existing rooftop units, ranging in age from 10 to 25 years old. Information suggested that these units had EERs ranging from 4.8 to 5.0, whereas the calculated EER of the new chiller under ice-making conditions was 8.0, and under chilled water conditions, 10.9. In this project, the efficiency of the ice storage system was found to be somewhat lower due to the need for pumps to circulate brine (ethylene glycol solution) between the chiller, the ice tanks, and the building.

System Design Approach

Project staff calculated that the chiller could provide 13.5 tons of refrigeration during the evening of a design day and at an ice-making water temperature of 26°F. This could produce over 100 ton-h of stored /"cool/" during the nighttime. From information provided by the vendor, staff determined that the tank could deliver no more than 140 ton-h of cooling for a 6-hour period without altering the glycol temperatures to the tank. Because the tank provides 140 ton-h of cooling during discharge, a subsequent charging process need only provide 140 ton-h of ice storage.

As ice is formed at night at a rate of 13.5 tons, it would take 10.4 hours for the system to make and store 140 ton-h. By discharging 140 ton-h in a 6 hour period, the average cooling delivered to the building is 23 tons. [Using a 20-ton chiller to provide 23 tons of cooling illustrates a key attribute of a cool storage system the capability for meeting a large cooling load with a smaller chiller.]

A DOE-2.1E model of the building was developed and calibrated using measured data on the total energy consumption of the building and selected components, including the five air handlers. The measured electrical energy consumption and simulated consumption for the building agreed to within five percent for the months June through September 1994—months when the ice storage system provides its best advantage. The model was used to predict the hourly loads of each of the rooftop units. This data file was used to determine the peak cooling load needed to complete the design of the system, and to provide a basis for Roofberg performance simulations as well as for non-storage alternatives.

The ice storage system started operating at the end of July 1995. Project staff collected 30-minute data on system temperatures, flowrates and energy consumption. They found that the energy consumed by the pumps tends to make a large difference in the energy consumption of the system. In this system, at least one pump was on around the clock to meet the continuous cooling load of a computer room. If this computer room had been provided with a separate cooling system, the pumping power could have been reduced and the ice storage system performance improved.

Alternative System Studies

Staff also examined systems likely to be evaluated by a building or facility manager in upgrading the building/'s cooling system. These cases included the baseline, existing rooftop scenario, the Roofberg application, improved Roofberg applications, and replacement of the existing rooftops with new rooftops. The purpose of these studies was to evaluate the Roofberg as one retrofit option for the building.

Case 1 (base case)—leaving the original rooftops in place. Although old, these units were continuing to cool zones in the building adequately; however, maintenance requirements for these units had become quite significant and annual maintenance costs high.

Case 2—the Roofberg system as designed, installed and tested. As per the design and to accommodate continuous cooling of the computer room, the pumps operated around the clock and a mixing valve with reduced port size was used.

Case 3—the Roofberg system with low pressure drop mixing valves that activate the pump only as needed for ice building or delivering cooling to the three main coils in the buildings. This eliminates the capability of providing continuous cooling to a computer room cooling coil.

Case 4—the Roofberg system with low pressure drop mixing valve and pump controls which together activate the pump only as needed for ice building or delivering cooling to the three main coils in the buildings. This eliminates the capability of providing continuous cooling to a computer room cooling coil.

Case 5—chiller only operation. Here, the ice storage system was removed, and the chiller operated as needed to provide cooling to the three main coils in the building. In this scenario, the three main coils were converted as for Roofberg, and coupled together so that the chiller could provide the cooling. This chiller used in this simulation was identical to the one for Roofberg.

Case 6—rooftop replacement. This case considers replacement of the old rooftops with new units of the same rated capacity.

Case 7—Again, a rooftop replacement except with smaller combined installed capacity.

The rooftop replacement options (Cases 6 and 7) and the original rooftop (Case 1) had the same impact on the total cooling load, as expected. With these rooftops, about 43% of the total cooling load occurred during the off-peak time (6:00 p.m. to 10:00 a.m.). With all of the options which used a chiller, including the Roofberg cases, the overall cooling load increased due to pump work.

Replacing the original rooftops with new ones (Cases 6 and 7) reduces the overall energy consumption by about 27%. This reduction is due to an improvement in energy efficiency. The chiller option (Case 5) had about the same total energy consumption as the downsized rooftop replacement option.

In the case of the Roofberg system as installed, the total energy consumption is about 20% higher than for the original rooftops. However, much of the energy penalty is due to extra pumping power requirements resulting from a higher pressure drop in the loop and from continuous pump operation. If it had not been necessary to run the pumps on a continuous basis to provide computer room cooling, the overall electrical energy consumption would have been 77% of the energy used by the original rooftop systems.

The principal advantage of the storage options is in electric load shifting from on-peak to off-peak hours. The Case 2 Roofberg system was able to reduce the on-peak energy use of the cooling system to 35% of the on-peak energy consumption of the baseline system. The benefit of load shifting is realized where there is a spread between on- and off-peak utility tariffs and where there are demand charges exacted for on-peak energy consumption.

Legal Notice | Download Centre | Site Map | Search
Copyright: Shanghai Tyacht Cooling System Co. Ltd.

2013   沪ICP 06051796