Of the 3,000 GWh fan energy consumed each year by large office buildings in California, only 30% is being used to supply air for ventilation purposes. The difference of 70% is used to transport the energy necessary to condition the space, by means of recirculated air. Hydronic distribution systems could fulfill the task of transporting the energy used for space conditioning by means of pumps at an energy level of approximately 5% of the otherwise necessary fan energy. This improvement alone significantly reduces the energy consumption, and the peak-power requirement of air-conditioning systems. Hydronic distribution systems have historically undergone changes in design and efficiency. In this paper we present a case study that demonstrates the savings potential of a high-performance Radiant Heating and Cooling System compared to an Air-and-Water System of the first generation.

The high-rise office building located at the west bank of Lake Merrit in downtown Oakland. Most of its 28 stories above ground accommodate offices, and are presently rented to several tenants. The innovative curved facades of the building, together with a large scale Hydronic Radiant Cooling air conditioning system, represented a modem design for an office building in the 1960s.

Thirty years and two oil-crises after its inauguration, there is anecdotal evidence that the specific cooling energy (W/m2) of the office building is significantly larger than that of buildings of comparable size located in downtown Oakland. This is surprising, since the building has approximately 60% of its 100,000 m2 ceiling area equipped with radiant cooling panels, and Hydronic Radiant Cooling Systems are known to require less cooling energy for operation than All-Air Systems that provide the same cooling effect. Faced not only with large energy bills, but also with a system that has been in operation for 30 years, the building management had the opportunity to study the heat transfer mechanisms in the building, and to examine the efficiency of a Hydronic Radiant Cooling System of the second generation, considered today's "state of the art".

A study of the heat transfers occurring in the office building enforced the belief that energy awareness was not an issue in the 1960s. As a general observation, a building having facades with minimum insulation, and single-glazed windows with aluminum frames, will experience extremely high heating and cooling loads. Furthermore, due to its location close to Lake Merrit, the reflection of solar radiation on the lake surface significantly increases the external loads on the south-east facade. Experimental data have shown interior glass surface temperatures of 50oC and cooling loads as much as 130 W/m2 floor area due to solar radiation on perimeter zones in the Fall cooling season. Fall is also the season when the 8.5 MW cooling capacity chillers work close to full load. Due to the limited cooling power of the first-generation radiant panel system installed (approximately 70 W/m2), and to the high cooling load of the building, a significant portion of the cooling load has to be removed by the air distribution system. Fall is therefore also the season when the supply and exhaust fans run at peak load, totaling approximately 9OOkW.

It is not too surprising that several energy consultants have tried their talents on this building. The energy savings measures proposed so far mostly dealt with changes in the plant (replace vapor compression chillers with absorption chillers), or with the replacement of the Air-and-Water System by an All-Air System. Variable speed fan motors have already been installed as a first measure to reduce fan power. However, the conversion to an All-Air System would eliminate some of the energy efficiency- and comfort-related features provided by Radiant Systems. Furthermore, the recirculated air associated with All-Air Systems contributes to the pollutant transport in the building. Gasses and particles which escape the filtering process are evenly distributed to all the zones served by a particular system.

Our project proposed a comparison of the already installed Hydronic Radiant System with a new, modem system of the same type. The idea was to show that modem Hydronic Radiant Cooling Systems are available today that can meet the cooling requirements of the building, while using less energy and providing a healthier indoor environment than All-Air Systems.

Due to the nature of the project, we had access to a study space for only four weeks. During this time we installed a new ceiling in the test room, and measurement equipment in both reference and test rooms; we monitored both systems; in the end we removed everything and reinstalled the old system. After the four week period, a new tenant had the interior of the whole floor rehabilitated.

SYSTEMS
Initially Installed System
The Burgess-Manning system installed in 23 stories of the office building consists of parallel steel pipes carrying the cooling fluid, and of structured aluminum panels attached to the pipes through steel clamps. Although the panel edges are bent to the radius of the steel pipe, the heat transfer from the coolant via the steel pipes to the aluminum panels, is limited due to lack of proper contact. This poor contact represents the cause for a much reduced cooling power output from this particular radiant cooling panel design.

The limited cooling power of the panel system leads to the necessity of having the air system remove part of the cooling load. In order to meet the load, air is supplied at 120C in a regime of 6 air changes per hour (ach). Furthermore, the air is supplied to the space through air diffusers in the ceiling, and exhausted through the ceiling panels. While passing through the cooling ceiling, the air is being re-cooled. The energy penalty of the installed system is increased by the fact that, in order to avoid occupant complaints about indoor air quality problems, the system operates with 100% outside air instead of using partially recirculated air.

Pilot Study
During our first site visit we realized that a plant retrofit would not address the real problem of this building, which consists of very high energy gains. By trying to compensate for these energy gains, instead of avoiding them, a plant retrofit would provide only limited savings. The most effective retrofit measure, the replacement of the existent windows with double-glazed, low-emittance windows with insulated frames, had unfortunately been ruled out because of cost. The window replacement would have reduced the cooling load to a level which the installed Hydronic Radiant Cooling System could have handled. However, because of the involved costs, it was clear that the facade would not be altered.

The second best choice was to change the design of the system application in the conditioned zone. This would provide the opportunity to reduce the amount of energy used for transporting the cooling fluid from the plant to the zone to be conditioned, and to simultaneously remove the cooling load more efficiently. A more efficient cooling panel system would then allow the downsizing of the air distribution system in the building, which would also result in lower equipment cost and reduced operating cost. Unfortunately, the local utility company does not have a rebate program for this kind of retrofit, and the local building codes do not even address this issue.

To prove that second-generation cooling panels are capable of removing high specific cooling loads, and thereby reduce the amount of air transported through the building, a test room in the building was equipped with a high-performance panel system. The panels installed were extruded aluminum panels, with the water channel integrated in the panel (see Figure below). This design limits the temperature difference between the cooling surface and the surface of the water channel, which helps to reduce the risk of condensation while maximizing the cooling power.

The supply air was reduced to the ASHRAE recommendations (ASHRAE Standard 62, 1989) and controlled to 10 L/s outside air per office occupant. The air was supplied through ceiling outlets close to the back wall, at low velocities, and at a temperature slightly lower than the room air temperature. The air flow patterns provided by this setup are similar to those of a displacement ventilation system. In order to remove some of the solar heat gain, the air was exhausted above the window, thus flushing the space between the glass and the interior blinds with room air. Ducted return runs were installed in order to channel the exhaust air. To avoid the influence of air exhausted from other spaces on the air exhausted from our test room, the air space above the test room was sealed from the common plenum of the floor by means of a gypsum board construction. For the pilot installation, a constant cooling agent mass flow with variable supply temperature was chosen to control the heat gain removal.

A room similar to the test room served as a reference room. The control of the hybrid system in the reference room was kept unchanged.

Cooling agent mass flow, supply and return temperatures, as well as air flows and temperatures, were monitored in both rooms. Infrared thermography was also applied to determine temperature distribution of the panel surfaces, and of the interior blinds surfaces. The equipment recorded the monitored parameters in the reference room first, followed by the same parameters in the test room.

Room dimensions for both the test room and the reference room are 25 m2 floor area, 2.7 m wall height, and 4.5 m facade width. The windows are 4.0 m wide and 2.0 m high, placed at 0.75 m above the floor level. Both rooms were located on the 23rd floor, on the south-east facade overlooking the lake.

Internal loads were simulated by 100 W incandescent lamps covered by metal cylinders. Six lamps provided a cooling load of 600 W, simulating two occupants and office equipment (see Figure below). Both ceiling types were operated at a design temperature difference between supply and return of 2K. Water flows and air flows were kept constant during the test period.

RESULTS
Peak-Power
The maximum cooling load per room, recorded during the first week of September 1993, was 3250 W (130 W/m2). The calculations performed in order to compare peak load handling by the two systems were based on an indoor air temperature of 24oC.

In the reference room, slightly more than half of the load was removed by the cooled ceiling, leaving the rest to the air system (400 m3/h at 12oC supply air temperature). The return air was exhausted from the spaces through a small number of ceiling panels which were not covered with insulation. This arrangement causes unnecessary re-cooling of the exhaust air. Our measurements enabled us to calculate that the loss due to cooling exhaust air was approximately 200 W.

In the test room, the air supplied at 18oC and 72 m3/h had only a cooling potential of 145 W. Due to the location of the air exhaust above the single-glass window behind the blinds, some of the warm air normally stagnant in this space was exhausted via the ducted exhaust. This captured an additional 145 W of the cooling load, warming the exhaust air to a temperature of approximately 30oC. The remaining 2950 W of the cooling load were removed by the water flowing through the aluminum panels of the cooled ceiling. This gave a cooling power for the ceiling of 296OW/22m2 = 134.5 W/m2 of active ceiling.

In a similar calculation for the whole building, the high performance radiant panel system allows the reduction of the supply and return air flows from 638,000 m3/h for the existing system to 108,000 m3/h. The cooling power of the air supply for the whole building corresponds to a specific heat removal rate of 43 W/m2. Based on measurements, the average power of the installed cooling ceilings for the whole building (facades with direct solar beam, facades with diffuse solar as well as internal offices and hallways) was assumed to be 2 49 W/m . For the peak -load scenario presented above, a comparison of the chiller capacities for the two radiant systems studied can be made. The following calculations are based on an outdoor air intake of 29oC and a humidity ratio of 13g/kg (ASHRAE Fundamentals, 1993).

In the first-generation system, 638,000 m3 /h of intake air are cooled to 10oC, and thereby dehumidified to 7g/kg (dh = 34 kJ/kg). Due to fan heat and conduction losses, the air is heated to 12oC before entering the room. The exhaust air leaves the room at a temperature of 24oC. The supply air temperature controls the moisture content of the indoor air to a level that the dew-point is always below the panel surface temperature.

The supply air temperature is also the moisture-controlling agent in the second-generation system. In general, the supply air temperature necessary to control the dew-point in the building determines the cold-deck temperature. Without significant humidity sources in the building, the dew-point of the supply air must be at least two degrees below the water supply temperature, in order to avoid condensation on the ceiling. The amount of air supplied is much reduced as compared to the system in the reference room (17%). Since displacement ventilation limits the range of acceptable supply air temperatures (approximately two to four degrees Kelvin below room air temperature), the air might have to be reheated before being supplied to the space.

The fan energy scales with the 3rd power of the air flow rate. Given that the air flow rate for the second generation system amounts to 17% of the rate of the current operation, the fan power would theoretically be reduced to 0.5% (4.5 kW) of the current value. This would, however, lead to pressures which do not allow any balancing of this system anymore. In case the air flow were reduced, but the pressure level were kept unchanged, the relationship between air flow and power requirement would be linear, thus reducing the power requirement to 17% (152 kW) of its current level of 900 kW.

CONCLUSIONS
Given the poor performance of the first-generation ceiling panels installed in the office building, a large fraction of the cooling load is presently removed by the ventilation system. In order to avoid occupant complaints, the air recirculation was eliminated, so the system supplies the building with outside air only. The large cooling load in this building should, however, not be removed by means of air flow, because of the high transport energy required for recirculating the air.

Our study shows that second-generation cooling panels, i.e. extruded aluminum panels with integrated water channels, perform far better than the first-generation panels originally installed. The study shows that the comfort level inside the building is improved by using radiant cooling more efficiently, supplying low-velocity air to the space, and exhausting the air above the window. Furthermore, panels with low resistance to heat transfer reduce the risk of condensation by allowing the cooling system to operate at a high coolant temperature level. The features of this system define it as a valuable candidate for energy-aware designers.

Home

This page was developed at IES by hef. Last changed on July 21, 1997
If you have any comments, please send us mail to the above address. Thanks.