Design issues for a/c split systems

In December 1997, the EU undertook to reduce emissions of greenhouse gases to 8% below the 1990 level by 2010. The UK Government has set an aspirational target of cutting emissions by 20%. This is a challenging target and calls for big improvements in the energy efficiency of building services.

With the increased use of IT and improvements in fabric insulation, comfort cooling is needed for a large part of the year in many UK buildings. The design of such comfort cooling systems should not only consider initial cost, but also energy consumption, CO₂, emissions and running costs.

Unfortuntely, the person who foots the bill for a comfort cooling system often has little or no say in the choice of equipment when it was installed. This is left to the building services engineer or air conditioning distributor/contractor who often selects a system based on initial cost and heuristic knowledge.

As a result, the equipment's country of origin rarely comes into the equation, which means that split-type comfort cooling systems manufactured by Far East conglomerates now dominate the UK market.

Sizing issues

Split air conditioning systems are rated using the standard JIS B 8616, which describes an air condition of 278C db/198C wb and ambient condition of 358C db/248C wb.

With these values the room air has a moisture content of 10·4 g/kg and a specific enthalpy of 53·6 kJ/kg. The UK equivalents are just 7 g/kg and 39 kJ/kg respectively.

Such humid conditions are often experienced in countries between latitudes 308N and 308S, where the latent cooling load can be up to 50% of the total cooling load. This explains why systems designed for these countries generally have low sensible heat ratios.

The UK, situated between 50 and 608N, has completely different cooling load profiles. In this case, sensible heat ratios typically range between 0·8 and 0·9.

Testing contemporary coolers

A typical contemporary comfort cooling system was tested by the author under the JIS conditions. While it was found to operate satisfactorily and generally in accordance with the manufacturer's published data, the coefficient of performance (cop) was found to be only 2·19 and the indoor unit air leaving temperature was 13·38C db with a sensible heat ratio of 0·54.

The same system was tested under typical UK indoor conditions, with considerably less moisture in the room air.

In these conditions a higher percentage of the mechanical refrigeration capacity is devoted to sensible cooling. As the air flow rate is fixed, the result is a much lower leaving air temperature and saturated evaporating temperature.

In this case, the temperature of the air leaving the indoor unit was -0·58C db – ideal for a coldroom, but a far cry from comfort cooling. The saturated evaporating temperature at the compressor was an astonishing -148C, with frost forming on the suction line and accumulator.

Good air conditioning practice dictates that the temperature differential between the room and supply air should be around 9K. Anything greater than 11K causes serious problems with air distribution and draughts.

The contemporary system tested under UK conditions had a massive temperature differential of 21·5K. This would make those unfortunate enough to be sitting near one of these units feel very uncomfortable.

This differential results in a drop in efficiency because the evaporating temperature is so low. The vapour compression is also much less efficient when evaporating temperatures are lower, due to the greater specific volume of vapour entering the compressor.

The low exit air temperature and low evaporating temperatures are caused by the condensing unit's inflexible refrigeration capacity. The indoor unit struggles to cope with a latent load that is much lower than the manufacturer anticipated.

Coefficient of performance

Under typical UK conditions, the highest possible Carnot coefficient of performance1 (cop) for a comfort cooling system is 73·75. This is very attractive as it implies that the system would yield 73·75 kW of cooling for an energy input of only 1 kW.

In practice, however, the cop is always much lower than this ideal value – largely because heat transfer occurs with finite temperature differences.

Refrigerant pipework in the system also reduces the cop, due to a high irreversibility in the suction line. In long pipe runs this can amount to as much as 20%, resulting in a 23% increase in CO₂ emissions, running costs, and energy consumption. Additional or larger cooling units would also be needed for a given cooling load.

The change in the saturated suction temperature resulting from the suction line pressure drop can be as high as 13K, producing an evaporating temperature of -178C at the compressor. This reduces the cop to around 2·0.

Good engineering practice dictates that the suction line pressure drop should correspond to no more than 2·0K (0·04 K/m). Table 1 shows the variation of coefficient of performance values for a typical comfort cooling system.

The evaporating temperature for comfort cooling systems should normally be from 2-48C to prevent freezing of condensate on the evaporator. Some manufacturers try to overcome this by incorporating frost prevention devices into the control circuit. However, this only causes the units to shut down.

Assessing actual performance

A cooling system specifically designed for UK use is different. The summary second law analysis below shows the coefficient of performance to be much higher: Compressor:

To(s2a - s1) = 296 x (1·81 -1·7661)m = 0·502 kW

Expansion device: To(s4 - s3)m = 296 x (1·1532 -1·13942)m = 0·158 kW

Condenser: To[(s3 - s2a) + (Qo/To)]m = 296 [(1·13942 - 1·810) + (353·32 - 142)/296]m = 0·495 kW

Evaporator: To[(s1- s4) + (Qr/Tr)]m = 296 [(1·7661 - 1·1532) + (142 - 310·41)/294]m = 0·458 kW

Total irreversibilities = 1·613 kW

Carnot cop = Tr/(To - Tr) = 295/(299 - 295) = 73·75

Carnot work = RE/cop = 6·5/73·75 = 0·088 kW Total work = It + Cw = 1·613 + 0·088 = 1·656 kW

cop (including fan power) = 6·5/2·02 = 3·22

Where:

cop is the coefficient of performance;

Cw is Carnot work;

It is total irreversibilities;

m is the mass flow rate;

Qo is the heat rejected to the atmosphere;

Qr is the heat absorbed in the evaporator;

RE is the refrigerating effect;

s is the specific entropy;

To is ambient temperature;

Tr is refrigerated space temperature.

The air flow rate for a temperature difference of 11K and a sensible cooling load of 5·2 kW is 0·39 m³/s. With an apparatus dewpoint (adp) of 8·18C, an evaporating temperature of 28C gives a reasonably sized evaporator.

By contrast, a contemporary system with the same cooling duty has an air flow rate of only 0·23 m³/s and an evaporating temperature of -148C (see figure 1 for state points.)

Design implications

Tests suggest that many contemporary comfort cooling systems are unsuitable for use in the UK. A mismatch between indoor and outdoor units creates the dual problems of low leaving temperatures and frost formation on the evaporator.

Root causes are insufficient air flow and evaporator surface area for UK conditions. This, together with losses caused by the interconnecting pipework, leads to a significant drop in efficiency and increases energy consumption, running costs and CO₂ emissions.

Hence specifiers should be very wary when selecting comfort cooling systems. The smallprint in technical manuals quotes a minimum indoor wet bulb temperature of 15·58C for guaranteed operation, which equates to a relative humidity of 55% at 218C db. This humidity is impossible to maintain in UK comfort cooled buildings without humidification.

Even with the wet bulb at 15·58C, the temperature differential of around 18K is too high. In effect, manufacturers are admitting that the units are unsuitable for UK conditions.

The problems are even more severe for applications which require year-round cooling. Here, the indoor relative humidity can drop to below 30% (12·88C wb at 218C db) during the winter months.

Unbelievable as it might seem, the majority of manufacturers do not know the sensible heat ratio of their systems when operating under UK conditions. The sensible heat ratio greatly influences energy consumption and the number of units required.

Indeed, at least two manufacturers were unable to give the author a cop for UK conditions. Specifiers should also be wary of the drop in performance caused by the interconnecting pipework.

As manufacturers' published data is simply not applicable for the climate, it is virtually impossible to select a comfort cooling system that is well suited to the UK.

By contrast, a properly configured comfort cooling system serving a medium-sized office block with an annual sensible cooling energy requirement of 525 GJ would yield savings in all areas.

Specifiers should question manufacturers' technical data and not simply select units from price lists, which sadly has become common practice within the industry. Consultants certainly cannot afford to rely on inaccurate advice from manufacturers and distributors.

The UK is clearly in urgent need of a UK-specific standard for comfort cooling systems.