Proposing an Equation to Evaluate Greenhouse Heat Transmission

Abstract

In a conventional greenhouse heating design, heat transmission through covering has been evaluated in proportion to the inside-outside temperature difference. The coefficient used in this proportional expression is eventually affected by both the weather conditions, e.g. wind speed, cloudiness, and the internal conditions, e.g. supplemental curtain layers. The heat transmission coefficient (h_t) customarily used, therefore, varies in a wide range according to those influencial factors. In order to take into consideration such variables in determining heat transmission, a new equation is proposed here.
q_t=1/1/(h_c+h_r)+1/h_i{T_in-T_ou+(1+β)R_n/2(h_c+h_r)} (12)
where, q_t; heat transmission, h_c, h_r; convective and radiative heat transfer coefficient at the greenhouse outside surface, respectively, h_i; overall heat transfer coefficient between inside air and external covering, T_in, T_ou; inside and outside air temperature, respectively, β; floor to surface area ratio (floor area/surface area), R_n; longwave net radiation on the open bare field.
The parameters h_c and h_r is estimated from outside wind speed and outside air temperature, respectively. The parameter h_i may be a function of covering conditions and should be determined experimentally. The third term in the braces corresponds to an equivalent temperature due to radiation.
The data observed in the greenhouses with three different covering types were used to estimate errors arising from eq. (12). They are a) PVC single covering, b) PVC single+PE single curtain and c) air inflated double covering with improved PE, infrared transmissivity of which is reduced to the same as EVA.
The coefficient h_i was first calculated from the experimental data. The h_i thus obtained showed little deviation according to outside weather conditions and almost constant values inherent in the covering types. Using these values for h_i, eq. (12) was evaluated and compared with the observed data. The comparison convinced us that the maximum errors in the estimation due to eq. (12) was within ±10% under various weather and internal covering conditions.