Modelling Low Energy Swimming Pools Adapted to Climate Change

Modelling Low Energy Swimming Pools adapted to Climate Change a

T. Kershawa, J. Fitzsimmonsb Centre for Energy and the Environment Rennes Drive, Exeter, EX4 4RN b Gale and Snowden Architects Exeter Bank Chambers 67 High Street, Exeter EX4 3DT

Corresponding Author: t.j.kershaw@exeter.ac.uk +44 (0)1392 724144

Abstract This paper compares several methods of calculating the evaporation of water from a swimming pool. Based upon guidance for the temperature and relative humidity in pool halls we calculate the rate of evaporation and present a methodology for incorporating the sensible and latent heat loads into a building thermal simulation software packages. We find that the energy required to maintain temperature and humidity levels in the pool hall is highly dependant on the fresh air supply. Being able to more accurately model the various plant loads in swimming pools will allow architects to arrive at lower energy designs. As a proof of concept we then use this method to compare different design features for a Passivhaus swimming pool to be constructed in Exeter. By considering the performance of the swimming pool in the current climate and under estimates of future climate change we are able to arrive at an optimal design. We find that without accounting for the different plant loads and their magnitudes within the thermal model incorrect assumptions about the suitability of different design features will be easy to make leading to energy wastage.

Introduction In the current economic climate there is an emphasis on reducing energy consumption and running costs. For buildings such as swimming pools a large part of the energy used will be to maintain the temperature of the pool water and the temperature and humidity of the pool hall, changing rooms and other areas. The Carbon Trust states that in a large hotel the pool alone can account for 10% of total energy costs [1]. The processes of heating / cooling and humidifying / dehumidifying are typically energy intensive and hence care must be taken when sizing and commissioning these systems to avoid wastage, it also means that swimming pools are an ideal candidates for the implementation of energy saving features and generation of renewable heat and energy. However, in mainstream dynamic thermal modelling software packages these loads cannot easily be accounted for, meaning that the true running costs may be hidden from the client and the impact of sustainable design features can only be estimated at the design stage. This paper reports on the design process for a novel new pool to be located in Exeter, the pool is to be of Passivhaus design [2], utilising a low loss design with energy saving features and has been designed with climate change resilience in mind. The new pool is to accommodate a 25m eight-lane county standard main pool, a 13m-learner pool and a leisure pool with water features, changing and staff facilities, reception, restaurant/cafĂŠ and offices. In addition a dry sports facility with two dance studios, a fitness studio and adequate changing facilities is to be included.

While the Passivhaus Planning Package will be used to show adherence to the Passivhaus design principles, a dynamic thermal modelling package is required to examine the interplay between weather, ventilation rates, various heat gains and internal environmental conditions.

In order to achieve the goals of occupant comfort and minimal energy usage, accurate modelling is necessary to ensure optimal design. However, thermal modelling software typically does not include bodies of water such as a swimming pool as a distinct entity and

hence the energy usage for maintaining water temperature, the rate of evaporation and the associated changes in humidity are not normally accounted for in the model. This paper presents a methodology for modelling the heat and energy transfers within a pool hall and applies them to the design of a low energy swimming pool.

Ventilating and heating pool halls can be rather complex and it is essential to manage these services correctly. The control of evaporation from the water surface is a function not normally encountered in the standard heating, ventilation and air-conditioning (HVAC) systems, and therefore can be misunderstood by designers and engineers. There is a direct link between the energy consumption of ventilation systems and evaporation of water vapour from the pool. The amount of heat in the pool, which is lost to evaporation, depends on the air conditions immediately above the pool. This energy, together with a small amount of heat loss through conduction and radiation, represents a major part of the energy exchange from the pool to the pool hall air. Controlling this is therefore the key to saving energy.

The HVAC system is normally the primary (or only) means of controlling the pool hall air quality, temperature and humidity. The need for controlling temperature and humidity is twofold, there is a need to prevent condensation and reduce the likelihood of corrosion damage.

There is also the need to produce comfortable environmental conditions for bathers, who would otherwise experience thermal discomfort due to reduced clothing levels and evaporative cooling from their skin. The HVAC system also plays a key role in removing contaminants such as Chlorine from the air.

There are various industry guidelines relating to the environmental conditions and fresh air circulation within a pool hall. The guidance documents are in good agreement about the internal air temperature (~30 °C, 1 °C above pool temperature) and a relative humidity (60% ±10 %) but are contradictory about ventilation rates. The Sport England ‘Swimming Pool Design Guidance Note’ [3] provides details for a fresh air change rates to be at 8-10 fresh air changes per hour (ac/h). This equates to a fresh air load of 17,500l/s for the proposed design.

This guidance seems misleading, as the actual fresh air exchange in litres per second will vary greatly depending upon the height of the pool hall for constant air changes per hour.

Since external air is typically cooler and drier than the internal air, if the fresh air change rate is too high this will lead to increased evaporation from the pool and increased heating and ventilation load and potentially humidification of the air to maintain occupant comfort.

Whereas the Good Practice Guide 219, ‘Energy efficiency in swimming pools’ [4] states: • 10 l/s per m2 of total pool hall area • 4 – 6 ac/h for standard use (8 – 10 for extensive water features) • Minimum 12 l/s per person • 100% fresh air operation should be available.

This guidance is somewhat confusing since it implies several different ventilation rates and the final statement indicates that this might not be 100% fresh air and that some can be recirculated, however, this is not stated explicitly. The general guideline of 10 l/s per m2 of pool hall typically equates to the 4–6 ac/h for many pool halls. Alternative guidance from CIBSE Guide B [5] references the good practice guide [4] for ventilation guidelines but further states that the ventilation rate may be reduced with occupancy to save energy.

Ventilation may also be recirculated to reduce fresh air supply to a minimum of 30%. The Carbon Trust [1] recommend similar levels again with ventilation of 10 l/s per m2 of pool hall typically equating to 4–6 ac/h. However, they recommend that the fresh air supply is supplied

by variable speed fans and controlled with a dew point sensor or a relative humidity sensor and fresh air is supplied primarily to control humidity and prevent condensation. Other requirements for fresh air are met by default. No minimum for fresh air supply is stated.

The different fresh air ventilation rates expressed in different guidance documents will likely result in confusion and lead to energy wastage if an inappropriate level is chosen. This variation in guidance may have arisen as a result of confusion over the fresh air supply rate and overall ventilation rate. The former required to control humidity, temperature and air quality, the latter to prevent condensation. As part of this study the impact of different fresh air supply rates will be investigated. Based upon the expected levels of occupancy in the pool hall (63 people) the typical fresh air load would be 750 l/s (~0.4 ac/h) at 12 l/s/person fresh air. A survey of other modern swimming pools using recirculatory ventilation systems reveals an air change rate of 4-5 air changes per hour with approximately 10-30% being fresh air.

Further a tour of a Passivhaus swimming pool in Germany found an air change rate of 1.5 ac/h with 30% being fresh air. It seems then that ~0.5 ac/h of fresh air is considered acceptable for controlling air quality, with sufficient air velocity provided by recirculation to prevent condensation. Since heating the fresh air supply is generally a major source of energy consumption for a pool building. Providing a simple heat exchange device, such as a plate heat exchanger can improve energy efficiency by reclaiming large amounts of lost energy from the exhaust air. Depending upon the type of heat exchanger utilised latent heat can also be reclaimed in addition to just sensible heat.

An optimum relative humidity is required in pool halls to ensure excessive energy is not used for dehumidifying the pool and also for adequate comfort levels. Setting the relative humidity too low will increase the rate of evaporation from the pool water and create an evaporative cooling effect on individuals when they are not in the pool. Setting the relative humidity too high can increase fabric condensation and corrosion of certain materials. Since there is likely to always be some evaporation from the pool even with the use of an insulated pool cover the humidification requirements of the pool hall should be met through natural evaporation so long as the fresh air change rate is not too high. By extension this means that infiltration should be as low as possible so that all ventilation is controllable. Evaporation of the pool water has a cooling effect on the pool and a latent gain on the pool hall. This is why a swimming pool must be continually heated even after it has reached temperature.

Evaporation of the pool water will be increased if the water is agitated by bathers or water features such as flumes or wave machines leading to an effectively increased surface area.

Thermal Modelling of Swimming Pools Thermal modelling of buildings is an important tool that can be used to predict energy usage for compliance purposes, but can also be used as a design tool. Thermal modelling can be effectively used to optimise the design of a building to minimise energy usage, maximise comfort and assess how a building will perform under different representations of weather and climate. There is however a problem when considering swimming pools, in that thermal models do not explicitly handle bodies of water and cannot estimate the latent and sensible plant loads. The aim of this report is describe a method for incorporating a representation of water evaporation into thermal models and to show how the design of a swimming pool building can be optimised for both the current and likely future climates using thermal modelling. The modelling software used in this study is the IES Virtual Environment [6], although the approach detailed here could be applied to other thermal modelling software packages. By calculating the rate of evaporation of water from the pool we can attribute a latent heat gain in W/m2, which can be included as a latent internal gain within the pool hall.

The sum of the sensible heat loss and the latent heat loss from the pool will approximate to the energy required to keep the pool water at the correct temperature. We then need to consider the energy required to heat fresh water to replace the evaporate. Figure 1 shows the dimensions of the main pool hall and the swimming pool that were investigated in this study.

The swimming pool is part of a larger complex but results presented are for the main pool only for clarity.

Figure 1 Illustration of the pool hall, showing representative dimensions.

Evaporation Rates There are several different methods for estimating the evaporation from a pool surface, allowing for different values to be estimated depending upon what variables are known.

According to the ASHRAE handbook [7] the rate of evaporation of water from a pool can be estimated according to the Carrier [8] equation, which when converted to metric units is given by: Wp =

A(0.087 + 0.07815V ) [ Pw − Pa ] Y

where, Wp is the rate of evaporation (kg/s), A is the area of the pool (m2), V is the air speed over the water surface (m/s), Y is the latent heat of water (kJ/kg), Pw is the saturation vapour pressure at the water surface temperature (kPa) and Pa is the saturation pressure at room dew point (kPa). Shah [9] noted however that there is some discrepancy with the use of this equation in the ASHRAE Handbooks. In early editions such as 1982 the equation is for unoccupied pools, however later versions (1999) the equation is for pools with normal activity. The 1987 version used in this report does not mention a level of activity for the equation. This change of use may be a result of reports that the Carrier equation over estimates pool evaporation [10]. ASHRAE now provides utilisation factors to modify the Carrier equation for the actual levels of use.

There are many factors that will increase evaporation from a pool and these can be estimated and used to modify the rate of evaporation given for an unoccupied pool. Shah [9] details several factors that will increase evaporation rates, these are as follows; waves on the water surface, wet-deck (surrounding tiled area) around the pool, wet bodies of pool occupants, spray caused by activity. These can be used to adjust the rate of evaporation for an unoccupied pool E0 to give an actual evaporation E by effectively increasing the total area over which evaporation can occur. Where,

ER =

E Apool + Awetdeck + Abodies + Awaves + Aspray . = E0 Apool

Thus the evaporation from an unoccupied pool is increased according to the ratio of the increased surface area (A) provided by waves, bodies, wet deck and spray. These areas can be estimated according to a pool utilisation factor Fu that is defined by: Fu =

Amax .

Apool N

Where Amax is the pool area per person at maximum occupancy and N is the number of pool occupants. Biasin and Krumme [11] gave values from German standards according to which Amax is almost constant at ~4.5 m2 per person for ordinary swimming pools. Using this utilisation factor Fu the different areas can be estimated as: Awetdeck = Fu Apool and Abodies = 0.3Fu Apool .

Aspray can typically be ignored under normal conditions and is only important in diving or sports pools [9]. Smith [12] estimated that the waves on the pool surface typically increase the surface area by ~20 %, with waves 150 mm high at 900 mm intervals. Thus, Awaves = 0.2Apool .

While this relationship is independent of pool occupancy it is only valid for Fu > 0. There are problems with these relationships in that for higher levels of occupancy Awetdeck can easily exceed available space. For the pool considered here Apool = 425 m2 (for a 25 m × 17 m pool) but the total area of the pool hall is only 750 m2. For the expected typical levels of occupancy (12m2/person in the pool hall) we get an utilisation factor Fu = 0.66. Thus we can see that Awetdeck for typical occupancy is approaching the total available space, we could cap Awetdeck at 325 m2, but this is still unrealistic as it is unlikely that all the space around the pool would be wet even if the pool were at maximum occupancy. Thus while these factors allow modification of an equation for the rate of evaporation from an unoccupied pool such as the Carrier equation [8] there are limitations.

There are several relationships relating the evaporation from a pool to utilisation factor and environmental variables such as those proposed by Shah [9] and Biasin [11]. The relationship we will use in this paper is that of Smith [14]. The work of Smith [12-14] is widely referenced in the literature and if converted to metric units and written in the notation used above [9] gives the following relationship for evaporation (kg/m2h): E=

(0.068 + 0.063Fu ) ΔP × 3600 .

Y

The ×3600 term is to convert from evaporation per second to per hour for consistency with the other relationships. E is the evaporation rate (kg/m2h), Fu is the utilisation factor and ∆P is the difference (in kPa) between the saturated vapour pressure at the water surface temperature (Pw) and the partial vapour pressure at room temperature and humidity (Pr).

These are given by the following relationships:

Pw = 6.112e

" 17.67×Tw % $ ' # Tw +243.5 &

" 17.67×Tr %

$ ' R and Pr = hum × 6.122e# Tr +243.5 & , 100

where Rhum is the relative humidity and Tw and Tr are the temperatures of the water surface and the room respectively. Note these equations produce pressures in mBar while ∆P is in kilopascals (kPa) (1 mBar = 100 Pa).

This relationship produces a feasible rate of evaporation and has been widely used. This relationship also has the benefit that links evaporation rate to the utilisation factor accounting for use but doesn’t require knowledge of the air speed over the water surface, which is hard to estimate at the design stage and is required to use the Carrier equation suggested by ASHRAE. For these reasons this is the relationship that will be used in later sections, however the methodology presented in this paper does not discriminate between this and other methods of calculating the rate of evaporation.

Calculations Table 1 shows the maximum pool water temperatures as stipulated by the Pool Water Treatment Advisory Group (PWTAG) [3]. These are maximum temperatures and pool operators may run temperatures 1-2 °C lower to save energy.

Recommended maximum pool water temperatures Competitive swimming / diving / fitness Recreational, adult pool Leisure pools Children’s swimming Babies, young children, disabled Hydrotherapy Spa pools

PWTAG 1999

PWTAG 2009

27 °C 28 °C 29 °C 29 °C 30 °C -

28 °C 29 °C 30 °C 31 °C 32 °C 35 °C 40 °C

Table 1 Maximum pool temperatures adapted from Sport England Guidance [3].

If we assume that the main pool (425m2) is operated at 28 °C to give a good mix between comfort and energy saving. Then according to the above guidance then the pool hall air temperature would be 29 °C with a relative humidity (Rhum) of between 50% and 70%. This gives values for Pw, Pr and ∆P of; Pw = 3.779 kPa Pr = Rhum/100 × 4.007 kPa ∴ 0.974 kPa ≤ ∆P ≤ 1.776 kPa (for 70% ≤ Rhum ≤ 50%).

The latent heat of vaporisation for water is 2260 kJ/ kg, thus, for an utilisation factor Fu = 0.66 the rate of evaporation given by Smith’s relationship [14] per m2 of pool area is: E = 0.309 kg/m2h or a latent heat loss of 194 W/m2 (for 50% Rhum).

E = 0.240 kg/m2h or a latent heat loss of 150 W/m2 (for 60% Rhum).

E = 0.170 kg/m2h or a latent heat loss of 106 W/m2 (for 70% Rhum).

As expected the rate of evaporation from the water surface varies with relative humidity, this has to be balanced with the energy required to heat incoming air. If we know the evaporative

heat loss from the pool we can estimate the total heating energy requirements for the pool.

The above values are both the heating energy required to maintain pool temperature and also the latent gain into the pool hall. By incorporating a representation of these processes into a thermal modelling package it is possible to estimate at the design stage the different plant loads and the influence the architecture can have on the energy consumption.

Implementation in Thermal Modelling Software To implement this methodology into a thermal modelling software package such as IES [6], we need to create the geometry and constructions to account for the latent and sensible heat gains. To account for the sensible heat gain we need to represent the body of water as a room beneath the pool hall maintained at a constant temperature, with the adjoining ceiling (the surface of the water) represented as a window with transmittance = 1, absorbance = 0, reflectance = 0, refractive index = 1 and IR emissivity = 1 (water is almost a perfect black body at these temperatures). This combination of values will allow sensible radiation emitted to pass from the pool to the pool hall unimpeded. In order to ensure that the window representing the surface of the water is at the correct temperature the construction needs to be thin and have a low surface resistance (a value of 0.01 m2W/K was used in this study). This will allow the up-ward facing surface of the window representing the surface of the pool water to reach the correct temperature. These values should be used for both the internal and external surfaces of the glazing representing the water surface. Evaporation from the pool surface was estimated using formulae for the evaporation of water as detailed previously and incorporated into the pool hall as a latent heat gain. In addition air exchange including infiltration was turned off between the pool zone and the pool hall. This means that heat can only be transferred from the pool either by conduction through the pool walls (to the earth) or by radiation or convection into the pool hall above. The geometry of the pool can be seen in figure 2, relevant values for the constructions are were used for the walls and floor of the pool, which were U-values of 0.35 W/m2K and 0.25 W/m2K respectively.

Figure 2 Illustration of the room representing the pool and relative values.

The room representing the pool is heated to 28 째C continuously and has no other internal gains and no air exchanges with the outside or other spaces. A domestic hot water load was added to the pool to account for the energy required to heat replacement water for that lost by evaporation. This equated to hot water requirements of 72.25 l/h, 102 l/h and 131.33 l/h for Rhum = 50%, 60% and 70% respectively, the water supply temperature was 10 째C and the water was heated to 28 째C. A latent heat gain in W/m2 was added to the pool hall

corresponding to rate of evaporation from the pool, this was adjusted from the values shown above to account for the fact that the pool hall has a larger area than the pool. The room is set to 29 °C, the humidity set to 50%, 60% or 70%, corresponding to which rate of evaporation is used. Both the pool hall air temperature set point and the latent heat gain are controlled with a modulating profile linked to occupancy, this follows the assumption that a cover will be used on the pool outside hours to save energy. This cover will effectively reduce latent heat gain into the pool hall to zero. If a cover were not used the modulating profile will need to account for evaporation from an unoccupied pool.

To explore the effect of different levels of Rhum humidity was controlled in the ranges of 5060%, 55-65% and 60-70%. These chosen to represent each of the evaporation rates calculated at Rhum = 50%, 60% and 70% respectively. These ranges were set so that the relative humidity did not move outside of the range set in guidance without resorting to excessive control [3-5].

Annual energy used to heat replacement water attributed to pool water evaporation 13.3 MWh, 10.3 MWh and 7.3 MWh for the three humidity ranges (in increasing order). No consideration has been given to pool water refresh rates and the energy required to heat that water, however, this could be accounted for by simply increasing the domestic hot water supply in the model and would be the same in each case.

The latent heat from the evaporation of pool water is included as a miscellaneous gain in the pool hall with 0 W/m2 sensible gain and a latent heat gain equal to the latent heat loss from the pool water given above multiplied by the ratio Apool / Apool hall. In this way we can use the thermal modelling software to account for all the loads attributed to running a swimming pool. The domestic hot water load represents the heating of replacement water, the sensible heat load for the pool accounts for heat loss via conduction and radiation, the latent gain into the pool hall is equivalent to the evaporative cooling load on the pool water and hence the heating load required to maintain the pool temperature. These loads can be allotted to a specific fuel types in the software to keep track of these loads and costs separately from space heating costs for the rest of the building if desired.

Performing dynamic thermal simulations of the building using weather files indicative of the Exeter climate [15] with the latent heat load included and relative humidity control in the main pool hall allows us to investigate the effect of different fresh air supply rates. We find that at higher fresh air supply rates moisture and relative humidity are at the lower limit of the allowable range while at lower fresh air supply rates values are at the higher limit, during occupied hours. This implies a change in operation of the pool hall from one of humidification at high fresh air supplies to one of dehumidification at lower fresh air supplies. To minimise energy usage it is generally advisable to have as wide a deadband as possible between control while still maintaining comfort. In this case it should reduce instances of latent conditioning in the pool hall if control of Rhum is relaxed to 60% Âą10%.

However, we find that even with a relaxed control on Rhum, fresh air supply rate still has a large impact on the latent conditioning of the pool hall as shown in figure 3. We see that as the fresh air supply is decreased we move from a situation of humidifying the internal environment to one of dehumidification. Ideally we want to find the fresh air supply rate that will require the least amount of energy to be expended to maintain comfort and control condensation.

Figure 3 Plot of latent space conditioning for the pool hall. For clarity data shown for a single week in June at different fresh air supply rates with Rhum controlled to 60% Âą10%. Positive values indicate humidification while negative values are dehumidification.

50-60% Rhum 55-65% Rhum 60-70% Rhum 50-70% Rhum

8 ac/h 1021 / 0.6 1315 / 0.1 1612 / 0 1059 / 0

4 ac/h 429 / 0.6 594 / 0.2 763 / 0 466 / 0

2 ac/h 140 / 1.4 235 / 0.2 338 / 0 173 / 0.1

1 ac/h 17 / 11 62 / 1 127 / 0.1 38 / 0.3

0.5 ac/h 0.7 / 60 3 / 15 28 / 1 1/8

0.25 ac/h 0.4 / 112 1 / 63 2 / 17 0.1 / 52

Table 2 Humidification / dehumidification energy in MWh for the pool hall at different fresh air supply rates and ranges of relative humidity.

We can see from figure 3 and table 2 that as the fresh air supply is decreased we move from a situation of humidifying the internal spaces to one of dehumidifying, which is more common.

This rises from the fact that the outside air is typically cooler and has a lower moisture content than the pool hall and that there is a latent gain from the evaporation of water from the pool surface. The shaded combinations of relative humidity control and fresh air supply rate indicate the lowest energy states in terms of latent conditioning. Since the lowest energy consumption occurs at different fresh air supplies for different relative humidities we can assume that the Carbon Trust [1] recommendation to use variable speed fans to deliver fresh air dependant upon either a dew point or relative humidity sensor will reduce energy usage yet further.

For simplicity here we have assumed that the pool cover is perfectly fitting and that there is no evaporation overnight. However, in practice this is unlikely to be the case and the small humidification load indicated may well be met by overnight evaporation. In theory by considering the rate of fresh air supply for the pool hall it should be possible to design a swimming pool that does not require either humidification or dehumidification. However, in practice it is likely that some dehumidification will still be needed. The key is to minimise these loads as much as possible to minimise energy usage.

The results show that allowing the relative humidity to settle around 65% in the pool hall can offer significant dehumidification energy savings which would result in lower fresh air change rates, reduced heating and lower fan energy requirements while offering reduced risk of thermal discomfort or condensation. The latter is also achieved by minimisation of thermal bridges in the construction as part of the Passivhaus design. A relative humidity of 65% would present a load of ~8.8 MWh/yr for heating water to replace pool evaporate.

Once we have a methodology for the thermal simulation of swimming pool buildings, we can start to investigate the effects of likely energy saving technologies and climate change adaptation measures not just on sensible loads but also on the latent loads. For this investigation the fresh air exchange rate was set to 0.5ac/h (extra air movement to prevent condensation is provided by recirculation), the building is of a Passivhaus design with a highly insulated envelope and a mechanical ventilation system with heat recovery (80% efficient). The pool is heated to 28°C, the pool hall is heated to 29°C with cooling occurring at 32°C, and a dehumidification set point of 65%. Larger fresh air change rates may be required to remove aggressive chemical such as chlorine, for this pool the aim is to use ceramic filters and UV treatment to limit chemical usage thus keeping fresh air loads to a minimum. Daylighting and views to the outside from the pool hall is provided by glazing on one façade and by roof windows facing the same direction.

Figure 4 Illustration of building model showing the location and size of glazing.

Simulations were performed with 4 design summer year type weather files for Exeter, one represents the current climate and the others represent the median estimate of future climate for different future time periods under a high emissions scenario (A1FI) [15-16].

Since potential sites for this swimming pool allow various orientations of the design, the first variable to be investigated is the orientation of the building and the predominant direction of the glazing (both on facades and rooflights) as shown in figure 4. Figure 5 shows the impact of changing the orientation of the building on the heating, cooling and dehumidification plant loads. Note that the values presented are the plant loads and not energy used and as such does not take into account the coefficients of performance of the various plants. We find that orientating the building South produces the lowest heating load while orientating North produces zero cooling load. However, for this design the cooling load is negligible compared to the heating load even when considering the effects of climate change up to the 2080s. The dehumidification load is greatest for the Northerly orientation likely as a result of lower

temperatures and hence an increased relative humidity for a given moisture content. The Southerly and Westerly orientations are similar for humidification loads with the Westerly orientation displaying marginally greater dehumidification loads, however the margin is significantly greater for the heating and cooling loads. As such we can conclude the that orientating the building with the majority of glazing facing South results in the lowest overall plant loads.

Figure 5 Comparison of heating, cooling and dehumidification (dehum) loads for the swimming pool versus predominant orientation of glazing for different time periods.

It can be seen from figure 5 that dehumidification loads will increase as we move into the future. This is likely a result of warmer external air temperatures allowing more moisture content in the air for a given relative humidity. This leads to an increase in the dehumidification load for the pool hall as internal temperature and relative humidity are not expected to change as they are linked to human thermal comfort and condensation risk.

Figure 6 shows the moisture content of the external air over the year for the current climate and in the 2080s. The increase in moisture content as a result of climate change may well require further consideration of fresh air change rates to control relative humidity as we move into the future.

Figure 6 Plot of the moisture content of the external air over the year for the current climate and the 2080s climate.

Several different scenarios were considered in an attempt to minimise plant loads further and to improve resilience of the swimming pool against future climate change, the Southerly orientation was used in all cases. The scenarios considered are providing shading from direct solar gains by introducing 1m shades above the roof windows (located above the pool hall) and on the glazed faรงade. In the base case there was only a 0.5m overhang above the roof windows and no shade on the faรงade (see figure 4). The next scenario considered was to consider reducing insulation in the building envelope to NCM 2010 levels (infiltration and thermal bridging were left at Passivhaus levels). The intention here was to look at how the reduced insulation levels will affect heating and cooling levels in the future, as more heat will be able to flow from outside to inside potentially allowing for even greater reductions in heating loads in the future. To avoid overheating this scenario was performed in conjunction with the 1m overhangs above the glazing. The result being that the building becomes a more conventional design and construction albeit with good air tightness and minimal thermal bridging. The final scenario is the reduction of the fresh air change rate during periods of lower occupancy. The occupancy and evaporation from the pool were set to reflect and occupancy of 1 person per 20m2 rather than 1 person per 12m2, the fresh air rate was reduced in line with this from 0.5ac/h to 0.3ac/h. A lower air change rate can be used due to the lower latent load due to decreased activity in the pool and hence less evaporation and there will be less CO2 produced from the occupants that requires removal. We can see from table 2 that reducing the fresh air change rate with constant occupancy and relative humidity set points that the dehumidification load is significantly increased. However, there may be savings to be made during periods of lower occupancy when the latent load is lower.

Figure 7 Plot of the heating, cooling and dehumidification (dehum) loads for different scenarios over different time periods.

Figure 7 shows the heating, cooling and dehumidification loads for the different scenarios considered for different time periods. We find that the addition of shading to the South facing faรงade and increasing the overhang on the roof windows reduces the cooling load to zero but increases the heating in all time periods by a greater amount. Upon further investigation there is still a heat load during the summer months even in the 2080s and the solar gains were reducing this heat load. The adoption of building regulations type constructions resulted in even greater heating and dehumidification loads, although cooling loads remained at zero. It seems clear then that for swimming pools at least increasing insulation in the building envelope is the optimum solution both now and going forward into the future. This is due to the large heat loads required to keep the pool hall in the correct temperature range and the increased cooling load associated with increased overheating frequency is small compared to the heating requirements. Even as external temperatures increase due to climate change the increased conduction through the building envelope due to decreased insulation levels is insufficient to balance the extra heat lost at other times. Finally we find that reducing the fresh air change rate during times of lower occupancy still leads to higher heating loads than the base case, this is likely due to the removal of the sensible (and latent) heat loads of the occupants. This appears to out weigh the increase in heating requirement from increased fresh air change rates, this is likely due to the use of the heat exchanger to pre-warm incoming air.

The cooler air due to fewer occupants increased the relative humidity and hence the dehumidification load is greater. It should be noted however that the increased heating load and dehumidification load are smaller than the increase to the dehumidification load

presented in table 2 for decreased fresh air at full occupancy. Furthermore if the air change rate is not altered in line with occupancy then the humidity in the pool hall will fall leading to increased evaporation from the pool itself and increased water heating as well as potential thermal discomfort for the occupants. Over prolonged periods there may be the need to humidify the air to maintain levels. Hence it seems clear that in order to minimise the plant loads of the swimming pool for every day use then there needs to be a mechanism such as a CO2 or relative humidity sensor for controlling the fresh air change rate with the occupancy level.

These simulations showed that there are pros and cons to many design strategies, but also highlighted areas for further investigation. The poor performance of the solar shades suggests that increasing the glazed area may reduce plant loads yet further. This is not an obvious solution as although the windows are triple glazed Passivhaus certified units we would be replacing highly insulated façade, there is also the issue of greatly increasing the overheating risk. Two additional scenarios were considered, both doubled of the glazed area on the South facing façade one included no overhang or shade as per the base case and the other included a reduced 0.5m overhang. It was found that the heating load was decreased increased glazed area but the cooling load increased by a similar amount. For the increased glazed area with no overhang the heating load was decreased by a 3-4 MWh/yr over the different time periods while the cooling load was increased by 3 MWh/yr by 2050 and 5 MWh/yr by 2080.

Including an overhang to compensate for the increased glazed area meant that heating load was only reduced by ~1.5MWh over all time periods and cooling load was increased by a similar amount over the base case. This implies that for this building increasing glazing has no benefit. However, this also implies that if increased glazed areas are required to increase daylighting levels within the pool hall or desired for further interaction with external spaces then this can be achieved with minimal impact on building plant loads by increasing glazed area with suitable shading.

The results show that the key energy load for the building is heating the pool water to maintain temperature and heating fresh water are the greatest energy loads followed by heating the pool hall to maintain a high air temperature. Simulations show that the pool hall will require heating throughout the year even under a 2080s climate change scenario. The energy load for this space heating outweighs the loads for cooling and dehumidification even in future climates. Therefore the overall strategy should be to minimise the pool evaporation and heating load where possible, this can be achieved by: • • •

•

•

Maximising solar gain to the building throughout the year by orientating to the south with optimum glazing rations.

Minimising heat loss from the building by including high insulation and high airtightness standards such as those recommended by the Passivhaus standard.

Maintaining relative humidity levels of around 65% and fresh air rates of 0.5 ac/h for normal use are the optimum in terms of minimising energy loads, using variable speed fans to alter fresh air supply rate to balance humidity and water evaporation with occupancy will result in lower energy use.

Including shading devices has a detrimental effect on the energy requirements for the buildings, as it will reduce the solar gain that can contribute to the heating load and therefore increase the overall energy load Significantly increasing the glazing levels will result in a neutral energy effect, as although the heating energy load will be reduced across the climate change scenarios, it will increase the cooling loads by an equal amount in 2050 and a greater amount by

•

•

2080. This study is useful if glazing areas are required to increase daylight levels or views out.

The use of efficient mechanical processes and heat recovery (sensible and latent where possible) will become even more prevalent in future swimming pool buildings in particular at reducing dehumidification loads and heating colder incoming fresh air load.

Finally, lowering the pool water temperature will decrease evaporation and reduce water-heating requirements. The fresh air supply rate may also be reduced to maintain relative humidity.

Summary We have devised a methodology for the incorporation of swimming pools into thermal modelling software. We have presented proof of concept by modelling an example swimming pool and examining heat loads for different levels of ventilation and humidity. The method accounts for latent and sensible heat losses from the pool water and the required energy to heat water and maintain humidity in the pool hall. This methodology has been used to assess the impact of different design scenarios on the energy loads of the pool building in the current climate and under estimates of climate change over this century.

Acknowledgements This work was supported by the Technology Strategy Board project “Swim4Exeter” under the Design for Future Climate stream.

References [1] Carbon Trust, CTG009, Swimming Pools: A deeper look at energy efficiency (2008).

[2] Passivhaus UK http://www.passivhaus.org.uk (accessed 18/7/2013) [3] IES, Integrated Environmental Solutions, www.iesve.com [4] ASHRAE handbook, HVAC Systems and Applications, 1987.

[5] Carrier W.H. “The temperature of evaporation” ASHVE Trans. 24 25-50 (1918).

[6] Shah M.M. “Prediction of evaporation from occupied indoor swimming pools” Energy and Buildings 35 707-713 (2003).

[7] Shah M.M. “Rate of evaporation from undisturbed water pools to quiet air: evaluation of available correlations” International Journal HVAC&R Research 8 125-131 (2002).

[8] Biasin K., Krumme W. “Die Wasserverdunstung in einem Innenschwimmbad” Electrowaerme International 32 A115-A129 (1974).

[9] Smith C.C., Jones R., Löf G. “energy Requirements and Potential Savings for Heated Indoor Swimming Pools” ASHRAE Trans. 99 864 (1993).

[10] Smith C.C., Löf G.O., Jones R. “Rates of Evaporation from Swimming Pools in Active Use” ASHRAE Trans. 104 514-523 (1999).

[11] Smith C.C., Löf G.O., Jones R. “Measurement and Analysis of Evaporation from an Inactive Outdoor Swimming Pool” Solar Energy 53 3-7 (1994).

[12] Sport England. “Swimming Pools, Updated Guidance for 2011” Design Guidance Note. February Revision 003 (2011).

[13] Action Energy. “Energy Efficiency in Swimming Pools – for Centre Managers and Operators” Good Practice Guide (GPG) 219 (2000). (www.actionenergy.org.uk).

[14] Chartered Institution of Building Services Engineers (CIBSE). “Heating, Ventilation, Air Conditioning and Refrigeration” Guide B (2005).

[15] Eames M., Kershaw T., Coley D. “On the creation of future probabilistic design weather years from UKCP09” Building Serv. Eng. Res. Technol. 32, 127–142, 2011.

Weather files available from: http://www.exeter.ac.uk/cee/prometheus/ (last viewed 03/09/2012).

[16] Kershaw T., Eames M., Coley D. “Assessing the risk of climate change for buildings: A comparison between multi-year and probabilistic reference year simulations� Building and Environment, 46 1303-1308, 2011.