TOON HERMAN

2-STAGE EVAPORATIVE COOLING 28 June 2012

TWO-STAGE EVAPORATIVE COOLING WHAT IS 2-STAGE EVAPORATIVE COOLING?

Figure 1. Direct (single-stage) evaporative cooling Everybody knows and understands direct evaporative cooling. Consider the figure on the left as an example: A wet pack with a circulating water system. (Pretoria altitude and typical conditions.) Evaporation of water into the surrounding air cools the air in contact with it. The amount of heat that is needed to evaporate the water is drawn from the air. This results in a drop of the dry-bulb temperature of the air. The dry-bulb temperature approaches the wetbulb temperature of the entering air. The air enthalpy remains unchanged. (Ignore the energy inputs from fan and pump.)

Figure 2. Consider now what happens if one puts a cooling coil in front: Overall, nothing much has changed: The total energy content of the air remains the same. The air is cooled in the coil, but exactly the same amount of energy is returned in the wet pack. But something interesting has happened: Look at the unit leaving conditions from top to bottom: The leaving enthalpy gradually drops from 83.4 to 49.9 kJ/kg and the dry-bulb temperature from 24.9 to 16.3 oC. The top 40% has a higher enthalpy than the entering air and has therefore absorbed heat. The bottom 60% has a lower enthalpy than the entering air and has been cooled. The top 40% can now be called the ‘cooling tower” and the air with its heat content (called ‘process air’) can be rejected. The bottom 60%, at an average dry-bulb temperature of 17.3 oC, has a useful cooling capacity and can be sent to the space (‘primary air’ or ‘product air’). One can also choose to go for a lower product air temperature at the expense of rejecting more of the original air quantity. For example, by only retaining the bottom 30%, one achieves a supply temperature of 16.5 oC. Note:

The above is a somewhat unusual way of looking at 2-stage evaporative cooling. The conventional story goes as follows: With the above ambient conditions, water is cooled in a cooling tower to 21.4 oC. This water is then used in a cooling coil to cool ambient air from 32.8oCdb/19.6oCwb to 23.9oCdb/16.9oCwb. (First – or indirect – stage of cooling.) From there, the air is then sent to a wet pack for adiabatic cooling to 17.6oC. (Second – or direct – stage of cooling.)

Two-Stage Evaporative Cooling

Page 1 of 10

TOON HERMAN

2-STAGE EVAPORATIVE COOLING 28 June 2012

A BRIEF HISTORY OF EVAPORATIVE COOLING During the 1980’s and 1990’s, multi-stage evaporative cooling has been mainly used in industrial and retail applications and also as a pre-cooling stage for conventional air-conditioning. Studies indicated that the system could also be suitable for the air-conditioning of purpose-built, specially designed, energy-efficient buildings and some designs were prepared. However, they were not implemented because, in the end, the lending institutions considered it safer to stay with console units. As far as institutional office buildings are concerned, the use of two-stage evaporative cooling was never seriously considered However, in the late 1990’s, a prospective tenant for a building in an office park stated his only requirements as 'lots of fresh air' and the option of being able to open doors and windows. This was the ideal opportunity to propose a two-stage evaporative cooling system. The system was much unsophisticated: There was no provision for return air (opening windows …), operation was only demanded when cooling was required, therefore the system was off during winter, and heating was provided separately. Control was very coarse: a couple of space thermostats, averaged, controlled the (two) evaporative cooling units. But the inside conditions were remarkably comfortable, with temperatures similar to conventional airconditioning and a freshness and airiness that only can be achieved with a high flushing rate of 100 % outside air. This however had the effect of raising the expectations of the users and very soon it was forgotten that the system was not much more than a glorified ventilation plant. lts performance was judged against that of a conventional central air-conditioning system costing three times as much. From then onwards, two-stage evaporative cooling has slowly moved into the commercial, institutional and government building market as a viable alternative to conventional central air-conditioning such as variable volume systems with chilled water plants. The system itself has evolved: It is now far more sophisticated, with proper supply and return air arrangement, variable volume air terminals with heating and individual room temperature control. In fact the only difference with its conventional air-conditioning cousin is that it uses no mechanical refrigeration for cooling, but gets its cooling ‘for free’ out of the ambient air. This also means that the capital cost has gone up and is now only marginally lower than the cost of a conventional air-conditioning system.

WHY EVAPORATIVE COOLING? Two-stage evaporative cooling was first introduced to save energy. Air-conditioning has always been perceived as being the biggest energy user in a building and cooling as accounting for the largest part of that energy use. If must therefore be of great benefit to get that cooling part ‘for free’. But air-conditioning is no longer the energy guzzler it used to be. In a modern well-designed building, airconditioning only accounts for about 20-25% of the total energy consumption (more or less on a par with the humble plug load – the new elephant in the room). And conventional mechanical cooling has also become more efficient. In a typical simulation exercise, cooling (the most efficient system: water cooled chillers with spray economisers) consumed 40% of the airconditioning energy, with fans and heating each 30%. Air-conditioning used 26% of the building’s energy. With 2-stage evaporative cooling for the same building, the total energy consumption goes down by about 10 kWh/m2 per annum (from 134 to 124) and air-conditioning now accounts for 20% of the total consumption. The breakdown is as follows: ‘free cooling’ 15%, fans 63% and heating 22%. The fans are now, by far, the biggest energy users. Another good reason for going to evaporative cooling is the simplicity of the system, its reliability and ease of maintenance. Although most systems now have sophisticated individual room temperature control, we are still not convinced that this is necessary. Ideally, it should be seen as part of the building, unobtrusive in the background, providing basic comfort control for normal office applications. At the same time, it should allow the architect to go for a more open design in contact with the landscaped environment around the building.

Two-Stage Evaporative Cooling

Page 2 of 10

TOON HERMAN

2-STAGE EVAPORATIVE COOLING 28 June 2012

But is it real air-conditioning? Air-conditioning is defined as “a system that provides an atmosphere with controlled temperature, humidity and purity at all times, regardless of weather conditions.” Based on this definition, 2-stage evaporative cooling does not comply. But neither do most of the current conventional air-conditioning systems. Maybe a more appropriate definition would be: “Systems designed to regulate ambient conditions within buildings for comfort or for industrial processes”. This definition does not specify that mechanical refrigeration must be used to create cooling, only that the system must deliver conditions suitable for the use of the building space. Conditions achievable with 2-stage evaporative cooling throughout the year are discussed below.

SYSTEM DESCRIPTION This is the unit most commonly used for comfort applications:

Fig.3

System Schematic – Standard Unit

The above configuration, with return air fan, return air plenum and mixing plenum is for use in a conventional Central Variable Volume air-conditioning system. The supply- and return air fans are variable volume (speed control), controlled by the supply air riser static pressure It is the intention to control the supply air temperature as close to constant as possible. The unit shown above is the "compact" unit: The cooling tower pack (top) and the direct evaporative cooling pack (bottom) form one continuous unit with a water distribution section at the top and a sump with

circulation pump at the bottom. The compact 2-stage evaporative cooling unit consists of the following: 1. 2. 3. 4. 5.

The filter bank covering both the coil intake (primary or product air) and the cooling tower intake (secondary or process air). The cooling coil: This is the first step of cooling (indirect evaporative cooling). The direct cooling evaporator pack with sump and circulation pump. On top of it, and forming part of it, is the cooling tower pack with the water distribution set. The cooling tower fan (axial). (Secondary air). The supply air fan (Primary air).

Two-Stage Evaporative Cooling

Page 3 of 10

TOON HERMAN

2-STAGE EVAPORATIVE COOLING 28 June 2012

The return air fan plus the plenums and dampers have been introduced to adapt the 2-stage evaporative cooling unit to comfort air-conditioning in a (large) central system where traditionally an all-air system would have been used, with chilled water as cooling source. It is the function of the air handling unit to deliver air at a constant temperature and a constant (riser) pressure. As long as the outside air drybulb temperature is above the supply air temperature setpoint (say 17oC), the unit runs on full outside air and the pump and cooling tower fans are switched or controlled to maintain the setpoint. When the outside air drybulb drops below the setpoint, the pump and cooling tower fans are off and return air is mixed with the outside air to achieve the required supply air temperature. A constant riser pressure is maintained by controlling the speed of the supply air fan. Note that there is no heater bank in this unit.

Psychrometrics and heat balances:

Fig.4

Psychrometrics – Standard Unit

Something interesting happens here: After absorbing 117 kW of cooling over the cooling coil (work it out), the primary air actually returns 43 kW of that in the humidifier pack (2nd stage), for a nett cooling capacity of 73 kW. (This is also exactly the amount of heat rejected by the cooling tower – as it should be.) There is a sensible cooling from 21.6 to 17.9 oC in the 2nd stage, but this is not an adiabatic cooling process: The air returns 43 kW of the 117 kW cooling from the cooling coil. The humidifier pack actually acts as a 2nd stage cooling tower: It keeps on cooling the water by another 3.5 degC. This brings the water temperature down to 17 oC and this increases the cooling coil capacity to 117 kW. But it is only borrowed: The air has to return a portion of it in the humidifier pack. Nett Room Sensible Cooling Capacity (NRSCC) With fan heat pick-up: 1.0 degC and room design temp: 23.5 oC NRSCC

= =

Two-Stage Evaporative Cooling

10.0 m3/s x 1.01 kg/m3 x (1.012 + 1.89*0.0144) x (23.5-18.9) degC 48.3 kW

Page 4 of 10

TOON HERMAN

2-STAGE EVAPORATIVE COOLING 28 June 2012

SUPPLY AIR TEMPERATURE The above examples indicate that a supply air temperature of around 17 – 18 oC is a realistic target under the specified ambient conditions. However, one has to look at real hourly full-year weather data. Looking, for example, at 2007 for Pretoria and Johannesburg and trying to control the supply air temperature at 17 oC, there will be a certain number of hours when this temperature can not be achieved. Number of hours (out of 3024 operating hours per year) that the design supply air temperature of 17.0 oC is exceeded. Table 1. Achievable Supply Air Temperatures - Statistics 2007

Pretoria

Number of operating hours: oC

Number of hours

> 17.0 > 17.5 oC > 18.0 oC > 18.5 oC > 19.0 oC > 19.5 oC > 20.0 oC

Johannesburg

3 024 hrs

3 024 hrs

403 hrs 13.3% 271 hrs 9.0% 166 hrs 5.5% 85 hrs 2.8% 43 hrs 1.4% 25 hrs 0.8% 18 hrs 0.6%

167 hrs 103 hrs 60 hrs 24 hrs 14 hrs 6 hrs 4 hrs

5.5% 3.4% 2.0% 0.8% 0.5% 0.2% 0.1%

It is clear that Johannesburg is ideally suited for 2-stage evaporative cooling, more so than Pretoria. Most of the interest comes however from Pretoria and from the above figures it would make sense to design a Pretoria building for 18.0 oC supply air and Johannesburg for 17,0 oC. However, this would mean that Pretoria would require 20% more air for the same load. This might not be necessary as the following simulation results will show. For some of the time, the supply temperature will be above 17.0 oC and this will affect the room temperature. Investigating this is part of the simulation process. ROOM TEMPERATURES ACHIEVED It is the intention that a 2-stage evaporative system achieves the same room temperatures a conventional air-conditioning. The following results are for a typical 4-storey building, of average thermal efficiency, with a system designed for an internal temperature of 23 oC with supply air leaving the unit at 17 oC (for both Pretoria and Johannesburg). The room temperature is controlled at 22.0 oC ± 1.0 degC. The following is an extract of the summary sheet of the temperature statistics for some of the rooms As can be seen, a supply temperature above 17.0 oC does not necessarily have a proportional effect on the room temperatures. Table 2. Room Temperature statistics – Typical Example 2007 ROOM:

Pretoria CB_NO

Johannesburg

CB_EA CB_SO CB_WE CB_INT

CB_NO CB_EA CB_SO CB_WE CB_INT

Max:

24.5

oC

23.5 oC

25.2 oC

24.5 oC

24.1 oC

24.0 oC 24.1 oC 24.4 oC 23.9 oC 23.6 oC

TR>23.0: TR>23.5: TR>24.0: TR>24.5: TR>25.0:

121 hrs 23 hrs 2 hrs 1 hrs 0 hrs

8 hrs 1 hrs 0 hrs 0 hrs 0 hrs

223 hrs 69 hrs 18 hrs 3 hrs 2 hrs

17 hrs 4 hrs 2 hrs 1 hrs 0 hrs

57 hrs 9 hrs 1 hrs 0 hrs 0 hrs

Discomfort Index: 25 (degC.hrs)

Two-Stage Evaporative Cooling

19 hrs 2 hrs 0 hrs 0 hrs 0 hrs

4 hrs 3 hrs 1 hrs 0 hrs 0 hrs

69 hrs 11 hrs 2 hrs 0 hrs 0 hrs

7 hrs 3 hrs 0 hrs 0 hrs 0 hrs

10 hrs 2 hrs 0 hrs 0 hrs 0 hrs

Discomfort Index: 6 (degC.hrs)

Page 5 of 10

TOON HERMAN

2-STAGE EVAPORATIVE COOLING 28 June 2012

The worst room, CB_SO exceeds the design temperature of 23 oC for (only) 7.4% of the time in Pretoria (2.3% in Johannesburg) and stands out from the rest of the building. It is clear that the façade and/or the system serving that room should to be looked at again. The term “Discomfort Index” requires some explanation: It is an area-weighted indicator of by how much and for how long the design temperature of 23.0 oC is exceeded on average throughout the building. For example, a Discomfort Index of 25 means that, on average, the room temperature exceeds 23.0 oC by 1 degC for 25 hours per year, or by 0.5 degC for 50 hours per year, or …etc. RELATIVE HUMIDITY Relative humidity is of concern with evaporative cooling and should be checked as part of the simulation process. For the same building as above, the statistics for the average relative humidity are as follows: Table 3. Relative Humidity - Statistics 2007 >60% >65% >70% >75% >80%

Pretoria 1 353 hrs 1 062 hrs 495 hrs 105 hrs 6 hrs

Johannesburg 45% 35% 16% 3% 0%

849 hrs 624 hrs 216 hrs 3 hrs 0 hrs

28% 21% 7% 0% 0%

It is generally believed that the relative humidity in office space should not be higher than 60% (ASHRAE says so …). However, we have experienced relative humidities of 75% and higher in an office environment and that was quite acceptable. (The temperature was 22 – 23 oC). From “An investigation of thermal comfort at high humidities” by Mark E. Fountain et al. the following extract: “McIntyre (1980) cites several studies showing that for operative temperatures within the comfort zone, differences in RH as disparate as 20% and 70% can be undetectable, let alone a source of discomfort.” Another potential problem with high humidity is mould growth. It is difficult to find relevant information on this subject. Most studies refer to winter conditions in the northern hemisphere and the growth of mould on surfaces on or near thermal bridges in the outside walls. A humidity level of 80% is mentioned but it is not clear if this applies to our conditions.

ENERGY and WATER CONSUMPTION Comparison with conventional air-conditioning, both for all-air variable volume. The most energy efficient conventional system uses a water-cooled chilled water plant with a spray-based economizer cycle. A more common conventional AC system employs air-cooled chillers and has no economizer. Table 4. Typical Energy and Water Consumption – Comparison Pta 2007

2-Stag Evap Cooling

Energy Fans Cooling Heating Total

Water cooled/Spray Econ 18.4 kWh/m2 p.a. 4.4 6.2 29.0 kWh/m2 p.a.

Difference Water

Two-Stage Evaporative Cooling

Conventional Air-conditioning

407 l/s-m2 p.a.

11.9 kWh/m2 p.a. 16.9 12.1 40.9 kWh/m2 p.a. 141% 11.9 kWh/m2 p.a. 348 l/s-m2 p.a

Air cooled/No Econ 11.9 kWh/m2 p.a. 40.4 12.1 64.4 kWh/m2 p.a. 222% 35.4 kWh/m2 p.a. 0 l/s-m2 p.a

Page 6 of 10

TOON HERMAN

2-STAGE EVAPORATIVE COOLING 28 June 2012

The systems simulated above are of the “orthodox” variable volume type, namely a constant supply air temperature, throughout the year, and for the total building, regardless of zone and orientation. This was the traditional configuration, when variable volume was first used, 40 – 50 years ago. A variable volume system however presents the designer with a dilemma: How to deal with the minimum air (required for ventilation and for the delivery of the heating) under low load conditions? In the early days, there was always a generous lighting load (up to 40 W/m 2) and even later, when the lighting dropped to 20 W/m 2, there was still sufficient load to deal with the minimum air at constant supply temperature. But lighting is now far more energy-efficient (as low as 5 W/m 2) and can no longer be counted on to disguise the minimum air reheat problem. Especially when day-light and occupation sensors are employed to reduce the lighting load even further. As can be seen in Table 4 heating energy consumption is substantial and a big chunk of that goes to minimum air reheat. Two-Stage evaporative cooling does a lot better than conventional because the supply air temperature is 4 – 5 degC higher and therefore requires less reheat. Designers have tried to reduce this waste of energy by zoning the building and adjusting the supply air temperature to suit the requirements of the zone. This of course assumes that all rooms in a zone have a similar load profile. It also involves the use of sophisticated control systems to somehow pick out the room in a particular zone with the highest cooling requirements and adjust the supply air temperature to suit that room. To fairly compare the energy performance of the two systems, we will ignore the heating consumption as this should be the same for all once the minimum air reheat is resolved. Table 4a. Typical Energy and Water Consumption – Comparison without Heating Pta 2007

2-Stag Evap Cooling

Energy Fans Cooling

18.4 kWh/m2 p.a. 4.4 22.8 kWh/m2 p.a.

Difference Water

407 l/s-m2 p.a.

Conventional Air-conditioning Water cooled/Spray Econ

Air cooled/No Econ

11.9 kWh/m2 p.a.

11.9 kWh/m2 p.a.

16.9

40.4 kWh/m2 p.a.

28.8 kWh/m2 p.a. 126% 6.0 kWh/m2 p.a. 348 l/s-m2 p.a.

52.3 kWh/m2 p.a. 230% 29.5 kWh/m2 p.a. 0 l/s-m2 p.a.

As can be seen, “free cooling” comes at a price. In the above example the price is a higher air quantity and therefore a higher fan energy consumption. But the difference is still 25% compared to the best conventional air-conditioning system. At the other end the spectrum, the air-cooled chiller system with no economizer (which is quite common) uses more than double the amount of energy. The differences (6 to 30 kWh/m 2 p.a.) may appear to be small when one looks at total building consumptions of 200 to 400 kWh/m2, but they become significant if one tries to achieve ultra-efficiency targets of 100 to 115 kWh/m2 p.a. for the total building consumption. OUTSIDE AIR In the case of the above simulation (Pretoria), the system runs on full outside air for 77% of the time. Return air is introduced when the outside air drybulb drops below 17 oC but the outside air supplied never drops below 11 l/s per person.

Two-Stage Evaporative Cooling

Page 7 of 10

TOON HERMAN

2-STAGE EVAPORATIVE COOLING 28 June 2012

OTHER APPLICATIONS 2-Temperature Units Going back to Fig. 2 and looking at the leaving drybulb:   

The top 40% is the heat rejection portion (cooling tower) The middle 30% delivers air at 18.1 oC and would be ideal for the interior areas. The bottom 30%, with air at 16.5 oC would serve the perimeter. Fig. 5 2-Temperature Unit – Psychrometrics

Fig. 6 2-Temperature Unit – System Schematic

Two-Stage Evaporative Cooling

Page 8 of 10

TOON HERMAN

2-STAGE EVAPORATIVE COOLING 28 June 2012

As a Cooling Tower (Air/Water Unit)

Fig.7 unit

Hybrid Cooling Tower / Air

The load could be a cooled slab and the primary air, at 18.7 oC, could be used to take care of the ventilation requirements, at the same time providing additional cooling. Nett Room Sensible Cooling Capacity (NRSCC) with fan heat pick-up: 1.0 degC and room design temp: 23.5 oC AIR: WATER:

10.0 m3/s x 1.01 kg/m3 x (1.012 + 1.89*0.0150) x (23.5-19.7) degC = 3.0 l/s x 4.19 kJ/kg-degC x (20.6-17.6) degC =

TOTAL:

40.1 kW 37.7 kW ------------77.8 kW

This is 67% more cooling than delivered by the standard (air only) unit. Indirect 2-Stage Evaporative Cooling Unit For people still not comfortable with the high humidity of the standard 2-stage evaporative cooling system, there is the 2-stage indirect version. (Fig. 8 and Fig. 9)

Fig.8 Indirect 2-Stage Evaporative Cooling Psychrometrics

Two-Stage Evaporative Cooling

Page 9 of 10

TOON HERMAN

2-STAGE EVAPORATIVE COOLING 28 June 2012

Note an interesting variation: The air leaving Pack 2 is not simply thrown away but turned around and used for the cooling tower (Pack 1). This reduces the total air quantity and its cooling capacity is even slightly better than that of the outside air.

Fig.9 Indirect 2-Stage Evaporative Cooling – System Schematic

No humidity is added to the supply air and the system remains basically a full outside air system, but the fan and pump energy is significantly increased so much so that most of the energy advantage compared to conventional air-conditioning is lost. And the same applies to capital cost. Maybe an option for applications where full outside air is required, but, in that case, heating will have to be introduced to maintain the required supply air temperature.

------------------------------------------------

Prepared by: A.F.E. HERMAN PrEng BSc(Eng) Toon Herman Associates Consulting Engineers [email protected]

Two-Stage Evaporative Cooling

Page 10 of 10