Volatile Organic Compounds (VOCs) constitute an important class of indoor air contaminants and they may cause adverse health effects for occupants in buildings. Indoor generated contaminants may be transferred between the supply and exhaust air streams of the building’s Heating, Ventilation and Air-conditioning (HVAC) system when air-to-air energy recovery devices are used. The run-around membrane energy exchanger (RAMEE) is a novel exchanger, which uses aqueous magnesium chloride (MgCl2) salt solution (34-35 wt%) as a liquid desiccant to transfer heat and moisture between remote supply and exhaust air streams. In the RAMEE, a gas-phase porous membrane is placed between the air stream and the liquid desiccant stream in each exchanger and the membrane prevents the salt solution from entering the air stream but still allows the transfer of water vapor through the semi-permeable membrane. In the RAMEE, VOCs may transfer between the exhaust and supply air streams due to (i) air leakage or (ii) due to dissolution of VOCs into the liquid desiccant in the exhaust exchanger and their subsequent evaporation into the air stream of the supply exchanger. These two transfer mechanisms were tested in the laboratory using two counter-cross-flow RAMEE prototypes (Prototype #4 and Prototype #6). Tests were conducted at different air and desiccant flow rates at AHRI standard summer and winter operating conditions. Sulfur hexafluoride (SF6) was used as a tracer gas to test air leakage and toluene (C7H8) and formaldehyde (HCHO) were used to test VOC dissolution and transfer. From an external source, a known concentration of VOC was injected into the exhaust air inlet stream and the transfer fraction of VOC to the supply air stream was calculated. This transfer fraction or Exhaust Air Transfer Ratio (EATR) defined by ANSI/ASHRAE Standard 84 (2012) at steady state conditions was used to quantify and compare the transfer fraction of contaminants in both prototypes. The uncertainty in the transfer fraction was calculated and all the uncertainty bounds were calculated for 95% confidence interval. The transfer fraction of sulfur hexafluoride was 0.02 +/- 3.6% for both prototypes tested, which means that the air leakage between the air streams is negligible. The transfer of toluene, which has a low solubility in water, was less than the uncertainty in the measurement. EATR* values for toluene were 2.3-3.4% and the uncertainties were 3.4-3.6%. The transfer of formaldehyde between the exhaust and the supply air streams was the highest and the EATR* values just exceeded the uncertainties in the EATR* measurement. The highest EATR* values for the transfer of formaldehyde in Prototype #4 and Prototype #6 were 6.4 +/- 3.6% and 5.3 +/- 3.6%, respectively. At steady state, the measured EATR* values for both prototypes were insensitive to changes in the air flow rate, the liquid desiccant flow rate, the latent effectiveness and the environmental conditions but time delays to reach steady state were significant. These results imply that there is a negligible transfer of contaminants due to air leakage between the air streams, a negligible transfer of low water soluble VOCs (such as toluene), but possibly a small detectable transfer of very water soluble VOCs (such as formaldehyde) between the exhaust and supply air streams of the RAMEE.

In this thesis, heat and moisture transfer between supply and exhaust air streams are investigated for a run-around system in which the coupling material is a desiccant coated solid that is transported between two exchangers. The finite difference method is used to solve the governing partial differential equations of the cross-flow heat exchangers in the supply and exhaust ducts. The outlet air properties are calculated for several inlet air operating conditions and desiccant properties. The accuracy of the heat transfer model is verified by comparing the simulations with well-known theoretical solutions for a single cross flow heat exchanger and a liquid coupled run-around system. The difference between the analytical predictions and the numerical model for sensible effectiveness for each exchanger and the run-around system were found to be less than 1% over a range of operating conditions. The model is also verified by modifying the boundary conditions to represent a counter flow energy wheel and comparing the calculated sensible, latent, and total effectiveness values with correlations in the literature. <p>Using the verified model for energy exchangers and the run-around energy recovery system, the sensible, latent and overall effectiveness are investigated in each exchanger and the run-around system during simultaneous heat and moisture transfer. The overall effectiveness of the run-around energy recovery system is dependent on the air flow rate, the solid desiccant flow rate, the desiccant properties, specific surface area, the size of each exchanger, and the inlet air operating conditions. The run-around system can achieve a high overall effectiveness when the flow rates and exchangers properties are properly chosen. Comparisons between the solid desiccant and salt solution run-around system effectiveness (Fan, 2005 and Fan et al, 2006) shows they are in good agreement. In a sensitivity study, the thickness of desiccant on the fibre is investigated in the solid run-around system. It was found that good performance is obtained with very thin desiccant coatings (1 or 2 micron). During the practical use of this system, a desiccant coated fibre could be inserted into very porous balls or cages that protect the desiccant coated fiber from mechanical wear. The performance sensitivity for this kind of run-around system is demonstrated.

In this thesis, heat and moisture transfer between supply and exhaust air streams are investigated for a run-around system in which the coupling material is a desiccant coated solid that is transported between two exchangers. The finite difference method is used to solve the governing partial differential equations of the cross-flow heat exchangers in the supply and exhaust ducts. The outlet air properties are calculated for several inlet air operating conditions and desiccant properties. The accuracy of the heat transfer model is verified by comparing the simulations with well-known theoretical solutions for a single cross flow heat exchanger and a liquid coupled run-around system. The difference between the analytical predictions and the numerical model for sensible effectiveness for each exchanger and the run-around system were found to be less than 1% over a range of operating conditions. The model is also verified by modifying the boundary conditions to represent a counter flow energy wheel and comparing the calculated sensible, latent, and total effectiveness values with correlations in the literature. <p>Using the verified model for energy exchangers and the run-around energy recovery system, the sensible, latent and overall effectiveness are investigated in each exchanger and the run-around system during simultaneous heat and moisture transfer. The overall effectiveness of the run-around energy recovery system is dependent on the air flow rate, the solid desiccant flow rate, the desiccant properties, specific surface area, the size of each exchanger, and the inlet air operating conditions. The run-around system can achieve a high overall effectiveness when the flow rates and exchangers properties are properly chosen. Comparisons between the solid desiccant and salt solution run-around system effectiveness (Fan, 2005 and Fan et al, 2006) shows they are in good agreement. In a sensitivity study, the thickness of desiccant on the fibre is investigated in the solid run-around system. It was found that good performance is obtained with very thin desiccant coatings (1 or 2 micron). During the practical use of this system, a desiccant coated fibre could be inserted into very porous balls or cages that protect the desiccant coated fiber from mechanical wear. The performance sensitivity for this kind of run-around system is demonstrated.

Heating, ventilation and air-conditioning (HVAC) systems are widely distributed over the world due to their capacity to adjust some local climate parameters, like temperature, relative humidity, cleanliness and distribution of the air until the desired levels verified in a hypothetical ideal climate. A review of buildings’ energy usage in developed countries shows that in the present this energy service is responsible for a portion of about 20% of the final energy usage on them, increasing up to 50% in hot-humid countries. In order to decrease this value, more and more different heat recovery systems have been developed and implemented over the last decades. Nowadays it is mandatory to install one of these units when the design conditions are above the limit values to avoid such components, what is possible to verify mostly in non-residential buildings. Each one of those units has its own performance and working characteristics that turns it more indicated to make part of a certain ventilation system in particular. Air-to-air energy recovery ventilation is based on the heat recovery transfer (latent and/or sensible) from the flow at high temperature to the flow at lower temperature, pre-warming the outdoor supply air (in the case of the winter). Therefore, it is important to understand in which concept those units have to be used and more important than that, how they work, helping to visualize their final effect on the HVAC system. The major aims of this study were to investigate the actual performance of the heat recovery units for LB20 and LB21 in building 99 at the University of Gävle and make some suggestions that could enhance their actual efficiency. Furthermore, the energy transfer rates associated to the heat recovery units were calculated in order to understand the impact of such components in the overall HVAC system as also the possible financial opportunity by making small improvements in the same units. To assess the system, values of temperature and flow (among others) were collected in the air stream and in the ethylene-glycol solution that works as heat transfer medium between air streams and is  enclosed in pipes that make part of the actual run-around heat recovery units. After some calculations, it was obtained that for the coldest day of measurements, the sensible effectiveness was 42% in LB20 and 47% in LB21, changing to 44% and 43% in the warmer day, respectively. The actual heat transfer representing the savings in the supply air stream is higher on the coldest day, with values of 46 kW in LB20 and 84 kW in LB21, justifying the existence of the heat recovery units even if those ones imply the use of hydraulic pumps to ensure the loop. The low values of efficiency have shown that both heat recovery units are working below the desired performance similarly to the pumps that make part of the same units.  This fact, together with the degradation of the units that is possible to observe in the local, indicates that a complete cleaning (followed by a change of the heat transfer medium) of the heat recovery units and a new adjustment of pumps and valves for the further changes, are necessary. By doing this, it is expected to see the year average sensible effectiveness increase to close to 45% in both units which will lead to a potential economic saving of around 41 000 SEK per year.

In this study, a numerical model is developed for determining coupled heat and moisture transfer in a run-around membrane energy exchanger (RAMEE) using two counter/cross flow exchangers and with a salt solution of MgCl2 as the coupling fluid. The counter/cross flow exchanger is a counter-flow exchanger with cross-flow inlet and outlet headers. The model is two-dimensional, steady-state and based on the physical principles of conservation of momentum, energy, and mass. The finite difference method is used in this model to discretize the governing equations.<p> The heat transfer model is validated with effectiveness correlations in the literature. It is shown that the difference between the numerical model and correlations is less than ¡À2% and ¡À2.5% for heat exchangers and run around heat exchangers (RAHE), respectively. The simultaneous heat and moisture transfer model is validated with data from another model and experiments. The inter-model comparison shows a difference of less than 1%. The experimental validation shows an average discrepancy of 1% to 17% between the experimental and numerical data for overall total effectiveness. At lower NTUs the numerical and experimental results show better agreement (e.g. within 1-4% at NTU=4).<p> The model for RAHE is used to develop new effectiveness correlations for the geometrically more complex counter/cross flow heat exchangers and RAHE systems. The correlations are developed to predict the response of the exchangers and overall system to the change of different design characteristics as it is determined by the model. Discrepancies between the simulated and correlated results are within ¡À2% for both the heat exchangers and the RAHE systems.<p> It is revealed by the model that the overall effectiveness of the counter/cross flow RAMEE depends on the entrance ratio (the ratio of the length of the inlet and outlet headers to the length of the exchanger, xi/x0), aspect ratio (the ratio of the height to the length of the exchanger, y0/x0), number of heat transfer units (NTU), heat capacity rate ratio (Cr*), number of mass transfer units (NTUm), and the mass flow rate ratio of pure salt in desiccant solution to dry air (m*). Beside these dimensionless parameters, the performance of the RAMEE system is affected by the liquid-air flow configuration and the operating inlet temperature and humidity.<p> This study concludes that the maximum effectiveness of the RAMEE system with two counter/cross flow exchangers occurs when NTU and NTUm are large (e.g. greater than 10). At any NTU, the overall effectiveness of the RAMEE system increases with Cr* until it reaches a maximum value when Cr*= . Increasing Cr* above causes the overall effectiveness to decrease slightly. Therefore, to achieve the maximum overall effectiveness of the system, Cr* must be close to . is a function of NTU and operating conditions e.g., with NTU=10, and under AHRI summer and winter operating conditions, respectively. The exchangers in the RAMEE system are needed to have a small aspect ratio (e.g. y0/x0<0.2) and small entrance ratio (e.g. xi/x0<0.1) to get the maximum overall effectiveness of a RAMEE system using two counter/cross flow exchangers. Such a RAMEE system has a total effectiveness 6% higher and 1.5% lower compared to the same cross-flow and counter-flow RAMEE, respectively (at NTU=10, Cr*¡Ö3, y0/x0=0.2 and xi/x0=0.1).

In this study, a novel counter-cross-flow run-around membrane energy exchanger (RAMEE) system was designed and tested in the laboratory. The RAMEE system consists of two (2) counter-cross-flow Liquid-to-Air Membrane Energy Exchangers (LAMEEs) to be located in the supply and exhaust air streams in the building Heating Ventilation and Air-Conditioning (HVAC) system. Inside each exchanger, a micro-porous membrane separates the air and liquid streams and allows transfer of the sensible and latent energy from the air stream to the liquid stream or vice-versa. The system exchanges sensible and latent energy between supply and exhaust air streams using a desiccant solution loop. The supply and exhaust air streams in the RAMEE can be located far apart from each other or adjacent to each other. The flexibility of non-adjacent ducting makes the RAMEE system a better alternative compared to available energy recovery systems for the retrofit of HVAC systems.<p> Two counter-cross-flow exchangers for the RAMEE system were designed based on an industry recommended standard which is to obtain a target overall system effectiveness of 65% for the RAMEE system at a face velocity of 2 m/s. The exchanger design was based on heat exchanger theory and counter-cross-flow design approach. An exchanger membrane surface aspect ratio (ratio of exchanger membrane surface height to exchanger length) of 1/9 and the desiccant solution entrance ratio (ratio of desiccant solution entrance length to exchanger length) of 1/24 were employed. Based on different heat transfer case studies, the energy transfer size of each exchanger was determined as 1800 mm x 200 mm x 86 mm. ProporeTM was used as the membrane material and Magnesium-Chloride solution was employed as the desiccant solution.<p> The RAMEE performance (sensible, latent and total effectiveness) was evaluated by testing the system in a run-around membrane energy exchanger test apparatus by varying the air stream and liquid solution-flow rates at standard summer and winter operating conditions. From the test data, the RAMEE effectiveness values were found to be sensitive to the air and solution flow rates. Maximum total effectiveness of 45% (summer condition) and 50% (winter condition) were measured at a face velocity ¡Ö 2 m/s. A comparison between the experimental and numerical results from the literature showed an average absolute discrepancy of 3% to 8% for the overall total system effectiveness. At a low number of heat transfer units, i.e. NTU = 4, the numerical and experimental results show agreement within 3% and at NTU = 12 the experimental data were 8% lower than the simulations. The counter-cross-flow RAMEE total system effectiveness were found to be 10% to 20% higher than those reported for a cross-flow RAMEE system by another researcher.<p> It is thought that discrepancies between experimental and predicted results (design and numerical effectiveness) may be due to the mal-distributed desiccant solution-flow, desiccant solution leakage, lower than expected water vapor permeability of the membrane, uncertainties in membrane properties (thickness and water vapor permeability) and heat loss/gain effects. Future research is needed to determine the exact cause of the discrepancies.

This thesis contains the numerically investigations of the performance of a run-around membrane energy exchanger (RAMEE) at different outdoor air conditions and the effects of non-uniform exchanger channels. The RAMEE is a new type of building ventilation air energy recovery system that allows heat and moisture to be transferred between isolated supply and exhaust air streams. Two liquid-to-air membrane energy exchangers (LAMEEs) are placed in the supply and exhaust air ducts and transfer heat and moisture between air and a circulating liquid desiccant that couples the two LAMEEs together. The ability of the system to transfer heat and moisture between isolated supply and exhaust ducts makes it appropriate for numerous HVAC applications (e.g., hospitals and building energy retrofits). <p> The performance of the RAMEE at different outdoor air conditions is shown to be highly variable due to the coupling of the heat and moisture transfer by the desiccant. This coupling allows the humidity ratio between the indoor and outdoor air to influence the heat transfer and the moisture transfer is influenced by the difference between the indoor and outdoor air temperatures. The coupling produces some complex RAMEE performance characteristics at some outdoor air conditions where the effectiveness values (i.e., sensible, latent, and total) were shown to be less than 0% or greater than 100%. Effectiveness and operating correlations are developed to describe these complex behaviours because existing correlations do not account for the coupling effects. The correlations can serve as design and operation tools for the RAMEE which do not require the use of an iterative computational numerical model.<p> Non-uniform exchanger channels are present in the RAMEE because of pressure differences between the air and solution channels which deform the membrane into the air channel. The non-uniform channels are analytically shown to create maldistributed fluid flows and variable heat and mass transfer coefficients. The combined effects of these two changes lead to a reduction in the RAMEE effectiveness, which increases as the size of the membrane deformation increases. The reduction in total effectiveness for an exchanger where the membrane has a peak deflection of 10% of the nominal air channel thickness operating at a NTU of 12 was shown to be 12.5%. These results of non-uniform exchanger channels agree with previously conducted experimental results.

In this thesis, a numerical model for coupled heat and moisture transfer in a run around membrane energy exchanger (RAMEE) with a liquid desiccant as a coupling fluid is developed. The numerical model is two dimensional, transient and is formulated using the finite difference method with an implicit time discretization. The model for the case of only heat transfer for a single heat exchanger is compared to an available analytical solution and good agreement is obtained. It is shown that the discrepancy between the numerical and theoretical dimensionless bulk outlet temperature of the fluids is less than 4% during the transient period. The model is also validated for the case of simultaneous heat and moisture transfer using experimental data measured during the laboratory testing of a RAMEE system. The results for both sensible and latent effectiveness showed satisfactory agreement at different operating conditions. However, there are some discrepancies between the simulation and the experimental data during the transient times. It is proposed that these discrepancies may be due to experimental flow distribution problems within the exchanger. The maximum average absolute differences between the measured and simulated transient effectivenesses were 7.5% and 10.3% for summer and winter operating conditions, respectively.<p> The transient response of the RAMEE system for step changes in the inlet supply air temperature and humidity ratio is presented using the numerical model. In addition, the system quasi steady state operating conditions are predicted as the system approaches its steady state operating condition. The effect of various dimensionless parameters on the transient response is predicted separately. These included: the number of heat transfer units, thermal capacity ratio, heat loss/gain ratio, storage volume ratio and the normalized initial salt solution concentration. It is shown that the initial salt solution concentration and the storage volume of the salt solution have significant impacts on the transient response of the system and the heat loss/gain rates from/to the circulated fluid flow can change the system quasi steady effectiveness substantially. The detailed study of the transient performance of the RAMEE is useful to determine the transient response time of the system under different practical situations.

This thesis contains the numerically investigations of the performance of a run-around membrane energy exchanger (RAMEE) at different outdoor air conditions and the effects of non-uniform exchanger channels. The RAMEE is a new type of building ventilation air energy recovery system that allows heat and moisture to be transferred between isolated supply and exhaust air streams. Two liquid-to-air membrane energy exchangers (LAMEEs) are placed in the supply and exhaust air ducts and transfer heat and moisture between air and a circulating liquid desiccant that couples the two LAMEEs together. The ability of the system to transfer heat and moisture between isolated supply and exhaust ducts makes it appropriate for numerous HVAC applications (e.g., hospitals and building energy retrofits). <p> The performance of the RAMEE at different outdoor air conditions is shown to be highly variable due to the coupling of the heat and moisture transfer by the desiccant. This coupling allows the humidity ratio between the indoor and outdoor air to influence the heat transfer and the moisture transfer is influenced by the difference between the indoor and outdoor air temperatures. The coupling produces some complex RAMEE performance characteristics at some outdoor air conditions where the effectiveness values (i.e., sensible, latent, and total) were shown to be less than 0% or greater than 100%. Effectiveness and operating correlations are developed to describe these complex behaviours because existing correlations do not account for the coupling effects. The correlations can serve as design and operation tools for the RAMEE which do not require the use of an iterative computational numerical model.<p> Non-uniform exchanger channels are present in the RAMEE because of pressure differences between the air and solution channels which deform the membrane into the air channel. The non-uniform channels are analytically shown to create maldistributed fluid flows and variable heat and mass transfer coefficients. The combined effects of these two changes lead to a reduction in the RAMEE effectiveness, which increases as the size of the membrane deformation increases. The reduction in total effectiveness for an exchanger where the membrane has a peak deflection of 10% of the nominal air channel thickness operating at a NTU of 12 was shown to be 12.5%. These results of non-uniform exchanger channels agree with previously conducted experimental results.

In this thesis, a numerical model for coupled heat and moisture transfer in a run around membrane energy exchanger (RAMEE) with a liquid desiccant as a coupling fluid is developed. The numerical model is two dimensional, transient and is formulated using the finite difference method with an implicit time discretization. The model for the case of only heat transfer for a single heat exchanger is compared to an available analytical solution and good agreement is obtained. It is shown that the discrepancy between the numerical and theoretical dimensionless bulk outlet temperature of the fluids is less than 4% during the transient period. The model is also validated for the case of simultaneous heat and moisture transfer using experimental data measured during the laboratory testing of a RAMEE system. The results for both sensible and latent effectiveness showed satisfactory agreement at different operating conditions. However, there are some discrepancies between the simulation and the experimental data during the transient times. It is proposed that these discrepancies may be due to experimental flow distribution problems within the exchanger. The maximum average absolute differences between the measured and simulated transient effectivenesses were 7.5% and 10.3% for summer and winter operating conditions, respectively.<p> The transient response of the RAMEE system for step changes in the inlet supply air temperature and humidity ratio is presented using the numerical model. In addition, the system quasi steady state operating conditions are predicted as the system approaches its steady state operating condition. The effect of various dimensionless parameters on the transient response is predicted separately. These included: the number of heat transfer units, thermal capacity ratio, heat loss/gain ratio, storage volume ratio and the normalized initial salt solution concentration. It is shown that the initial salt solution concentration and the storage volume of the salt solution have significant impacts on the transient response of the system and the heat loss/gain rates from/to the circulated fluid flow can change the system quasi steady effectiveness substantially. The detailed study of the transient performance of the RAMEE is useful to determine the transient response time of the system under different practical situations.