Tas study takes a close look at chilled beams
Published: 05 September, 2013
With most chilled-beam installations being for ‘informed’ clients, the Chilled Beams & Ceilings Association is working towards increasing the number of informed people. Its latest move is a detailed simulation of the performance of chilled beams. Ken Sharpe looks at the details.
For many years one of the principal arguments for using active and passive chilled beams rather than fan-coil units for air conditioning in buildings was their lower energy consumption. There were two major reasons for this lower energy consumption, both arising from the higher chilled-water temperatures used for beams compared with fan-coil units. Higher flow/return temperatures improve the efficiency of the chiller. In addition, those higher flow-return temperatures can permit the use of free cooling using ambient air and other cooling sources such as ground water or water from a lake or river.
What has been missing is a quantitative analysis of the relative energy consumption of chilled beams and fan-coil units. That situation has now been rectified with a detailed hour-by-hour building simulation carried out by EDSL using its Tas software at the instigation of the Chilled Beams & Ceilings Association, one of the specialist groups of FETA (Federation of Environmental Trade Associations). Until this study there was no chilled-beam product model to simulate and for users to specify, and all software providers treated a chilled beam as a fan-coil unit without a fan.
|An EDSL Tas simulation compared the energy required by air conditioning these four buildings using active chilled beams, passive chilled beams and fan-coil units.|
Andrew Jackson, chairman of the CBCA, explains that the work was carried as part of a continuing programme to promote a wider and deeper understanding of chilled beams and ceilings. That programme has already seen the publication of ‘An introduction to chilled beams and ceilings’ (available for download from the FETA website www.feta.co.uk, but you will have to hunt around a bit). The EDSL Tas analysis is also available to download, and a life-cycle analysis will become available in January 2014.
The typically detailed Tas analysis based on four types of building in London and Birmingham (see diagram). Homing in on the technical fact sheet produced by Andrew Gaskell, technical co-ordinator for the Chilled Beams & Ceilings Association within FETA, shows an annual energy cost savings of 17% for passive chilled beams compared with the VAV fan-coil system modelled and 22% for active chilled beams.
Presenting the simulation at a recent event for industry stakeholders, Dr Alan Jones, managing director of EDSL, was quite clear where the main energy savings came from. ‘It’s all about chilled-water supply temperature and the scope for free cooling using dry air coolers. Fans have only a secondary effect.’
The benefit from free cooling was more marked in Birmingham than in London.
The Tas simulation considered three air-conditioning systems in four buildings. Two had floor plates of 1750 m2 and were four and eight storeys high. The other two had floor plates of 3500 m2 and were four and eight storeys high.
Three air-conditioning systems were modelled.
• VAV fan-coil units with EC motors.
• Active chilled beams.
• Passive chilled beams (95% convection and 5% radiant).
Dr Alan Jones explains, ‘The fundamental difference in the specifications was the chilled-water supply and return temperatures, which were 6/12°C for the fan-coil units and 14/17°C for the active chilled beam. We were asked to use these temperatures to represent the most widely used practice and to create a typical base line, as both systems can elevate their chilled-water flow temperature from their respective base lines.
|Note the use of chilled beams from Frenger Systems in this IT room in the Hallwood Building of Nottingham University.|
‘The improvements in chiller COP and amount of cooling from dry air coolers offered by the high chilled-water temperatures is and was the fundamental different in energy consumption for the alternative systems. The VAV fan-coil terminal fans are so efficient that they have a secondary effect on relative energy use.’
Those flow/return temperatures for fan-coil units have caused a stir in the industry from FETA’s own Fan Coil Unit Group. The reason for the high return temperature from chilled beams is to keep the air-off temperature above the dewpoint and avoid condensation. Fan-coil units have, of course, a means for handling condensate.
John Lightfoot, chairman of the FCU group, comments, that comparing 6/12°C for FCU systems with 14/17°C for active and passive chilled beams makes the comparison very biased towards chilled beams. He suggests that to produce a fair comparison, 14/17°C could have been used for all three systems, adding, ‘Designers can of course use 14/17°C for fan-coil systems if they wise to further enhance a very flexible and efficient system.’
The Chilled Beams & Ceilings Associations acknowledges that higher flow/return temperatures can indeed be used with FCUs. However, it remarks that the commercially sensible maximum chilled-water flow temperature that a VAV fan-coil system could be raised to would be around 10°C, as anything above this would involve dehumidifying the air at the air-handling unit, which would add cost to the system and further compound the cooling performance of the FCU system given that fan-coil units are designed to condense and that forms part of their cooling capacity.
CBCA also comments that the smaller difference between the chilled water temperature and room temperature would reduce the performance of the FCUs, so more would be required, increasing the capital cost.
The Fan Coil Unit Group has indicated its intention to issue its own paper defining all the advantages that a fan-coil system can offer the whole supply chain within an office building/refurbishment programme. We’ll let you know when it happens.
In the meantime, the study that exists is that on active and passive chilled beams.
Alan Jones stresses how the simulation was set up to avoid bias towards one system or the other. Generic performance data agreed between all parties was used, even though Tas plant and control models can use laboratory results for manufacturers’ performance data. The central plant was common to all system configurations and comprised a chiller, dry air cooling and free cooling combination for heat rejection and a gas-fired boiler for heating. Conditioning of fresh air for temperature and humidity consisted of a typical VRF heat-recovery direct-expansion circuit.
There was much primary discussion about specific fan power (SFP) and the effect that the various systems would have on SFP at the air-handling unit. There was also some debate on other system configurations, and the agreed scope of work was a compromise between all parties. Alan Jones explains, ‘The use of AHU SFPs across all systems was one such case. The active-beam SFP would have at worst case an additional 100 Pa to cope with [to induce room air, as active chilled beams do not have secondary fans], adding a little more than 0.1 W/l/s [to the 2.1 W/l/s used]. This was modelled and had no significant impact on energy use.’
Another aspect of the study was to run fans in the VAV terminals at minimum fresh air, 20%, which would probably cause dumping of cool air. An additional simulation turning down to 60% increased energy used by terminal fans by 11.5%, representing an overall increase in fan-energy use of 2.2%. Alan Jones comments, ‘60% flow drops energy use by about 80%. There is little benefit in turning down further.’
|Tas plant and control models can use laboratory results for manufacturers’ performance data.|
The lower fresh-air rate would have narrowed the gap between active chilled beams and fan-coil units by only 1% from 22% to 21%.
Finally, why is there five percentage points difference in the energy consumption of a passive chilled-beam system versus an active system. Back to Andrew Gaskell, who explains that the passive beam’s displacement ventilation system requires a higher fresh-air supply temperature (to meet occupant comfort) than that of the active system. That higher temperature results in more energy being used on the fresh-air re-heat DX circuit and also less airside cooling being available. At certain times of the year, therefore, active systems will have more free cooling airside than a passive system, and the passive system will have to make up any shortfall of airside cooling using the chiller — increasing its energy use.
Putting the study into context, Michael Ainley, a member of the CBCA summarises: ‘The evidence regarding how much more energy efficient chilled-beam technology is compared to other forms of air conditioning is long overdue and comes at a time when an uplift for commercial buildings of 9% is required for the new Part L, which is due to be implemented in October 2013 and come into force by April 2014.’