THERMAL ENVIRONMENT MODELING AND OPTIMIZATION OF GREENHOUSE IN COLD REGIONS

Thermal simulation models for the time-dependent heating requirement of greenhouses are very important for the evaluation of various energy-saving technologies, and energy-efficient design of greenhouses based on local climates. A quasi-steady state thermal model “GREENHEAT” was developed using the programing language MATLAB for simulation heating requirement in conventional greenhouses. The model could predict the hourly heating requirement based on the input of hourly weather data, indoor environmental parameters, and physical and thermal properties of greenhouse building materials. The model was validated with measured data from a commercial greenhouse located in Saskatoon, Canada, and the monthly average error in prediction was found to be less than 5.0%. This study also reviewed various energy-saving technologies used in greenhouses in cold climate, and the GREENHEAT model allowed selections of commonly used ones in the simulation. The GREENHEAT model was used for evaluating the impact of various geometrical parameters on the heating requirement of the single span and multiple-span conventional greenhouses located in Saskatoon. Results showed that the east-west oriented gable roof greenhouse could be more energy-efficient for the multi-span gutter connected greenhouse whereas quonset shape as a free-standing single span greenhouse. The large span width could be beneficial for the single-span greenhouses, but the impact of increased span width could be negligible on the heating demand of multi-span greenhouses. The model was also used for an economic feasibility analysis of year-round vegetable production (tomato, cucumber, and pepper) in northern Saskatchewan, and tomato was found to be the most economical vegetable as compared to the cucumber and pepper. Another heating simulation model CSGHEAT was developed to estimate of the supplemental heating requirement of mono-slope Chinese-style solar greenhouses (CSGs). This model is also a quasi-steady state thermal model using the programming language MATLAB, and it can simulates the hourly heating requirement of CSGs. The model was validated with experimental data from a CSG located in Winnipeg, Manitoba. The average error for prediction of the hourly heating requirement was found to be less than 8.7%. The model sensitivities to various geometrical and thermal parameters were studied. The results indicated that the thermal properties of cover, thermal blanket, and parameter insulation were the most important design parameters in CSGs. Finally, the heating requirement in CSGs was modeled using TRNSYS simulation tool, and the predictions were compared with that of CSGHEAT. The result indicated that TRNSYS had serious limitations for modeling of greenhouse thermal environment, thereby high uncertainties could occur, thus was not suitable for greenhouse simulation