Dehumidification Technology Evaluation and Moisture Balance Modelling for Greenhouse Humidity Control

Abstract

Excessively high relative humidity (RH) occurred in the greenhouses almost all year around. Various methods of dehumidification are available for greenhouses. To find a feasible method for greenhouse dehumidification, three methods including air-to-air heat exchangers, exhaust ventilation system, as well as the mechanical refrigeration dehumidification were compared in a tomato greenhouse in the cold region of Canadian Prairies. The experiment results showed that dehumidification by the exhaust fan system was the most cost-effective method with the lowest capital and maintenance cost. However, similar to the heat exchangers, the exhaust fan system is only effective during cold and mild seasons, and not during warm weather conditions. Even though the mechanical refrigeration dehumidifiers consumed the highest amount of electrical energy thus resulting in the highest cost, they were effective in controlling the indoor moisture year-round due to their independence from outside air conditions. Mechanical refrigeration is recommended for summer dehumidification which is only needed at night and early morning before ventilation cooling starts. Both methods could be used during different seasons to achieve good moisture control year-round. A moisture balance model for simulating the greenhouse indoor RH and air water vapor partial pressure was developed. The model, named HumidMod, takes plant evapotranspiration as the main moisture source of greenhouse air, which is calculated by a modified Penman-Monteith evapotranspiration model. Condensation on the greenhouse inner cover surface as one of the moisture sinks or sources is calculated by two statistical models developed in a Venlo-type plastic greenhouse. Ventilation or infiltration is estimated as a function of the indoor solar radiation. In the model, the indoor RH and water vapor partial pressure can be directly calculated as a function of the indoor and outdoor air conditions, as well as the plant and greenhouse characteristics. The model was validated by comparing predictions with measured data in a tomato greenhouse, which had a commercial-grade refrigeration dehumidifier for humidity control. The mean absolute uncertainty between the predicted and measured results was about 6.9% for both RH and water vapor partial pressure. The coefficient of determinations were 0.59 and 0.75 for RH and water vapor partial pressure, respectively. A good agreement was found between the predicted and measured results with root mean square error of 5.6% for RH and 0.144 kPa for water vapor partial pressure. This model provides a reliable tool for the estimation of dehumidification requirement inside a greenhouse to achieve a desired humidity level. Sensitivity analysis of this model to several important input parameters was also conducted in three different seasons: cold winter (January), mild season (April), and summer season (July). The results indicate that the input parameters including the indoor air temperature, incoming solar radiation, air exchange rate, as well as plant leaf area index have a significant influence on the model output so should be decided carefully.