3.1. Methodology
Various factors influence greenhouse energy performance, including shape, orientation, thermal properties of the envelope, ventilation methods, and equipment choices. The transparent envelope plays a crucial role in blocking or absorbing heat while providing the necessary light to support crop growth. These characteristics directly impact energy consumption and crop productivity.
Polyethylene is the most commonly used material for greenhouse envelopes due to its economic advantages In the Republic of Korea, more than 98% of greenhouses are constructed using plastic film, primarily polyethylene (PE). However, this material has limitations in thermal insulation, which leads to excessive energy consumption for heating in winter and cooling in summer. To address these inefficiencies, recent developments in greenhouse envelope materials have introduced alternatives such as polymethyl methacrylate (PMMA), polycarbonate panels, and glass. These materials provide enhanced thermal performance, but they also pose challenges in terms of installation cost and light transmittance. In response to the growing demand for energy-efficient greenhouses, research into low-emissivity (low-E) coated glass has gained attention, particularly in the Netherlands. Low-E coatings selectively control infrared radiation while allowing visible light to pass through, making them an effective solution for reducing heat loss in winter and minimizing solar heat gain in summer. This study evaluates the energy-saving potential of different transparent envelope materials and proposes an innovative rotatable low-E-coated glass system to optimize greenhouse energy performance.
To achieve this objective, a comprehensive methodology was adopted, encompassing data collection, energy simulation, regression analysis, and system development. First, a dataset was compiled containing 206 different glass configurations, including single-pane, double-pane, and triple-pane glass, with and without low-E coatings. The key thermal and optical properties analyzed in this study include the U-value (thermal transmittance), Solar Heat Gain Coefficient (SHGC), and visible light transmittance. These parameters serve as independent variables influencing the heating and cooling loads of the greenhouse.
The energy performance of different envelope configurations was assessed using DesignBuilder 7.3, a simulation tool based on the EnergyPlus engine. DesignBuilder enables the calculation of heating and cooling loads under various environmental conditions by considering factors such as external climate, greenhouse envelope properties, and ventilation rates. Additionally, the WINDOW 8.0 software was utilized to analyze the thermal and optical properties of glass and to calculate surface temperatures of different glazing materials.
A multiple regression analysis was conducted to quantify the relationship between glazing properties and greenhouse heating/cooling loads. The U-value and SHGC were designated as independent variables, while heating and cooling loads served as dependent variables.
Based on the regression analysis and simulation results, an innovative rotatable low-E-coated glass system was developed. The proposed system allows seasonal adjustment of the low-E coating orientation, optimizing greenhouse energy efficiency throughout the year. In winter, the low-E coating faces inward, reflecting infrared radiation back into the greenhouse and minimizing heat loss. In summer, the low-E coating faces outward, reducing solar heat accumulation and improving cooling efficiency. This system maintains high visible light transmittance while enhancing seasonal adaptability. To verify the feasibility of the proposed system, a patent was filed in the Republic of Korea under the title “Greenhouse with Rotary Panels” (Patent Number: 10-2024-0092864).
This methodological approach provides a comprehensive evaluation of greenhouse envelope performance and introduces a scalable, energy-efficient solution for sustainable agriculture. The findings contribute to the advancement of energy-efficient greenhouse designs by integrating simulation-based analysis, statistical modeling, and practical system development. The overall research process is outlined in
Figure 1, illustrating the step-by-step approach from data analysis to the development of an energy-efficient greenhouse envelope system.
3.3. Heating and Cooling Load Based on the Thermal Performance of the Envelope
Generally, for buildings, superior insulation performance of the envelope, that is, a lower U-value, enhances both heating and cooling efficiency. However, in transparent envelopes, the U-value and shading coefficient complicate the heat exchange. Therefore, optimizing the greenhouse envelope requires investigating trends in cooling and heating loads based on the U-value and solar heat gain performance.
Because greenhouses typically have five surfaces composed of transparent materials, careful consideration of the thermal properties of the envelope during summer and winter is essential. Unlike opaque materials, the transparent ones must account for solar radiation, making the U-value and solar heat gain coefficient (SHGC) particularly important. In some cases, transparent materials with high insulation performance (low U-values) may fail to dissipate absorbed solar heat near the glass surface through conduction or radiation, leading to heat accumulation. Consequently, high-insulation glass may inadvertently increase the cooling load inside the greenhouse.
Table 1,
Figure 3 shows the thermal performance results of the glass systems analyzed using the WINDOW 8.0 simulation program. During winter, the indoor surface temperature increased as insulation improved. Specifically, the application of low-E glass resulted in an indoor surface temperature increase of at least 4 °C. However, this enhancement led to an unintended rise in the indoor surface temperature of the glass. Although the SHGC decreased by 11% with low-E glass (as seen in cases 2 and 3 in
Figure 3), the improved insulation increased the indoor surface temperature by 2.4 °C. Therefore, while lower U-values were advantageous for heating during winter, they had a detrimental effect in summer by increasing indoor surface temperature.
To identify clearer relationships, simulations and statistical analyses were conducted to examine cooling and heating load consumption trends based on the greenhouse envelope. Two independent variables, the U-value and shading coefficient, were considered to determine the thermal performance of the transparent envelope, whereas the dependent variables were the cooling and heating loads. The building model used was the same as described in
Section 3.2; however, to minimize the influence of other variables, the model assumed ground installation.
The simulation used the DesignBuilder library, analyzing 206 different glass types across six categories, including single-pane, double-pane, and triple-pane configurations, with and without low-E coatings.
Figure 4 illustrates the distribution of thermal transmittance (U-value) and Solar Heat Gain Coefficient (SHGC) for these six categories. The U-values range from 0.5 to 5.0 W/m
2K, while the SHGC values are distributed between 0.1 and 1.0.
The thermal performance characteristics demonstrate distinct patterns based on glass configuration. A clear trend of improving insulation performance is observed in the progression from single-pane to double-pane to triple-pane, each exhibiting clearly defined U-value ranges. The SHGC values, however, show broad distribution patterns across all glass types, indicating that these values are not primarily dependent on the number of panes. Rather, the analysis reveals that SHGC values are more significantly influenced by factors such as glass color and the iron content. Nevertheless, a trend of decreasing maximum SHGC values is observed as the number of glass panes increases.
The heating and cooling load patterns were analyzed for 206 different glass configurations using DesignBuilder simulations.
Figure 5 and
Figure 6 present bubble chart analyses of heating and cooling loads in relation to thermal transmittance (U-value) and the Solar Heat Gain Coefficient (SHGC).
For heating loads, a distinct inverse correlation is observed with improved insulation performance (decreasing U-values). While the SHGC also influences heating loads, its impact leads to only marginal increases as values decrease.
By contrast, the cooling load patterns demonstrate inverse trends. Cooling loads rise significantly with higher SHGC values due to increased solar heat gain. Additionally, improved insulation (lower U-values) leads to a moderate increase in cooling loads, as absorbed solar radiation cannot dissipate effectively through conduction. In the Republic of Korea’s temperate climate, heating loads are approximately three times higher than cooling loads. Consequently, the annual heating and cooling load patterns predominantly follow heating load trends related to the U-value and SHGC. This relationship is further explored in subsequent sections through detailed statistical analysis.
Multiple regression analysis was conducted to examine the relationship between glass thermal properties and heating/cooling loads. Thermal transmittance (U-value) and the Solar Heat Gain Coefficient (SHGC) were designated as independent variables, while heating and cooling loads served as dependent variables. Prior to the regression analysis, potential multicollinearity between the independent variables was evaluated. The Variance Inflation Factor (VIF) was 3.69 for both variables, confirming their suitability for inclusion in the analysis.
The regression analysis for heating loads demonstrated high model reliability, with an R-squared value of 0.958 and an adjusted R-squared value of 0.957, indicating that the U-value and SHGC account for approximately 95.7% of the variance in heating loads. This high correlation is attributed to the greenhouse’s exclusively transparent envelope, which eliminates interference from opaque elements. Equation (1) represents the regression equation for heating load per unit area as a function of the U-value and SHGC. Consistent with the bubble chart analysis, the heating load is positively correlated with the U-value and negatively with the SHGC. The regression coefficients are −60.466 for the U-value and 16.893 for the SHGC. However, given that U-values range from 0.5 to 6.0 and SHGCs range from 0.1 to 1.0, the coefficients alone do not determine their relative impact on heating loads. A
t-test was conducted to evaluate the significance of each independent variable. In the heating load model, the U-value exhibited a t-value of 63.41, indicating strong significance, with a
p-value of 9.14 × 10
−136 (
p < 0.05). Similarly, the SHGC demonstrated statistical significance, with a t-value of −38.21 and a
p-value of 1.16 × 10
−94 (
p < 0.05) (
Table 2). While both variables contribute significantly to the model, the
t-test results indicate that the U-value is a more robust predictor of heating loads than the SHGC.
Figure 7 presents the multiple regression analysis results of the heating load in relation to the U-value and SHGC, along with residual distribution plots for reliability assessment.
Residual analysis demonstrated that all data points were normally distributed around zero, with no systematic bias. The residuals exhibit symmetrical patterns with no discernible skewness, indicating independence from predictor variables. This randomness in distribution validates the model’s predictive reliability.
The cooling load regression model yielded an R-squared value of 0.761 and an adjusted R-squared value of 0.759, explaining approximately 76.1% of the variance. Although lower than that of the heating load model, these values indicate reliable explanatory power (R-squared > 0.7). The weaker correlation suggests that factors beyond envelope thermal performance, such as greenhouse orientation and external climate conditions, exert greater influence on cooling loads.
Unlike heating loads, summer solar radiation intake and output mechanisms exhibit inverse correlations, with the cooling load negatively correlated with the U-value and positively correlated with the SHGC. The regression coefficients are −2.215 for the U-value and 48.763 for the SHGC (Equation (2)). However, as with heating loads, direct comparisons based solely on coefficient magnitudes are limited due to differences in the distribution ranges of the U-value and SHGC. Notably, the U-value has a substantially weaker effect on cooling loads than on heating loads (heating load U-value coefficient: 16.893; cooling load coefficient: −2.215). While the SHGC for cooling loads is somewhat lower than for heating loads, it remains influential (heating load SHGC: −60.446; cooling load coefficient: 48.763).
T-test results for the cooling load model show a t-value of −6.856 (
p = 8.35 × 10
−11,
p < 0.05) for the U-value and a t-value of 12.97 (
p = 5.30 × 10
−65,
p < 0.001) for the SHGC (
Table 3). While both variables demonstrate statistical significance in explaining cooling loads, contrary to the heating load model, the SHGC emerges as the dominant predictor.
Figure 8 presents the multiple regression analysis results for the cooling load in relation to the U-value and SHGC, along with residual distribution plots for model reliability assessment. Residual analysis confirms that the data meet normality assumptions, exhibiting predominantly symmetrical patterns with residuals centered around zero, further validating model reliability.
Analysis of annual heating and cooling loads confirmed the regression model’s reliability, with R
2 = 0.918 and Adjusted R
2 = 0.918. The overall trends reflect the characteristic pattern of heating loads being approximately three times higher than cooling loads annually. The results demonstrate that total energy loads decrease with improved insulation performance (lower U-values) and increased solar heat gain (higher SHGC values).
T-tests further validate the significance of both independent variables in explaining the regression model (Equation (3),
Table 4).
The distribution plots of predicted values and residuals exhibit symmetrical patterns consistent with heating and cooling load analyses. The absence of skewness or systematic patterns confirms the independence of residuals from predictor variables (
Figure 9).
The thermal characteristics of greenhouse envelopes present opposing requirements for cooling and heating loads. Heating load reduction benefits from low U-values and high SHGC values, with U-value reduction proving more effective. Conversely, cooling load reduction favors high U-values and low SHGC values, with SHGC control showing substantially greater effectiveness than U-value adjustments.
In temperate climates requiring both heating and cooling, adaptable transparent envelopes that respond to seasonal changes represent the optimal solution. However, where such adaptability is not feasible, envelope selection should prioritize controlling the predominant load type. In the Republic of Korea, where heating loads are approximately three times higher than cooling loads, optimizing the envelope for heating load reduction is the most effective approach.
3.4. Development of a Low-E-Coated Greenhouse
As demonstrated in
Section 3.3, temperate climates with distinct seasons require significantly different greenhouse envelope performance in summer and winter. To address this issue, thermochromic or electrochromic glass can be applied to vary insulation performance seasonally. However, these technologies only regulate the SHGC, rendering them insufficient as stand-alone solutions.
Typically, seasonal demands are managed through shading devices or by applying powder-type shading agents to the ceiling in summer and removing them in the fall. While this method controls solar radiation, it does not regulate insulation. Additionally, shading screens can obstruct visible light necessary for plant growth, reducing productivity.
Therefore, this study optimizes the thermal and optical properties of low-E coatings through physically controlling their thermal characteristics. Low-E-coated glass, featuring a nanoscale low-E coating, maintains high transmittance in the visible spectrum while selectively controlling long-wave infrared radiation.
In architecture, silver is commonly used as a low-E coating for insulation, typically applied in multilayers to prevent oxidation. This type of coating is well-suited for greenhouses, as it transmits visible light required for plant growth while regulating long-wave radiation for insulation. However, silver-coated low-E glass in double glazing presents challenges in cost-effectiveness and requires thicker structural frames to support the additional weight, limiting its greenhouse applications.
To address these limitations, this study explored tin oxide (SnO₂) as a coating material. While tin oxide offers lower thermal performance than silver, it provides superior durability, remaining stable even when exposed to air. This durability allows tin oxide-coated glass to function as a single-pane glazing system, eliminating the need for double glazing. When used as a single pane, the U-value and SHGC of low-E glass change depending on the orientation of the coating. If the coating faces inward, it reflects long-wave infrared radiation back inside, enhancing insulation. If oriented outward, this effect diminishes, yielding insulation performance (U-value) comparable to uncoated glass.
By utilizing the long-wave infrared radiation properties of low-E glass, greenhouse loads can be effectively controlled. In winter, an inward-facing coating reflects infrared radiation into the interior, minimizing heat loss and maintaining insulation. In summer, an outward-facing coating allows accumulated infrared radiation near the glass surface to dissipate outside, reducing cooling loads.
These characteristics enable optimal seasonal envelope control by adjusting the orientation of the coating. Greenhouses offer various openable designs, such as pivot-style rotating systems.
Figure 10 illustrates a rotating frame design that adjusts the low-E coating’s direction according to the season, along with thermal performance results analyzed using the WINDOW 8.0 program. Depending on the orientation of the low-E coating, the U-value varied by 37%, the SHGC fluctuated by 3%, while visible light transmittance remained unchanged.
These findings provide valuable insights into indoor environmental conditions based on glass system thermal performance, offering critical data for greenhouse envelope design. With greenhouses having five transparent surfaces excluding the floor, heating and cooling loads are highly sensitive to the thermal and optical properties of the envelope. Therefore, selecting appropriate materials is crucial to addressing distinct summer and winter requirements.