Elsevier

Solar Energy

Volume 177, 1 January 2019, Pages 595-603
Solar Energy

Predicting the water production of a solar seawater greenhouse desalination unit using multi-layer perceptron model

https://doi.org/10.1016/j.solener.2018.11.059Get rights and content

Highlights

  • An ANN model to predict the produced fresh water in the seawater greenhouse.

  • The Levenberg-Marquardt has been selected as the best training algorithm.

  • 161.6 m3/day of fresh water can be produced in this designed seawater greenhouse.

  • Solar energy and roof transparency have important effects on the water production.

Abstract

A solar seawater greenhouse is a type of desalination plant that uses solar energy and seawater to humidify the interior of the greenhouse and produce fresh water from the humid air (humidification-dehumidification process). The produced water is used both for irrigating agricultural crops and for drinking. Many parameters affect the performance of seawater greenhouses. The present study employed an artificial neural network to examine the effective parameters of the greenhouse on the fresh water production such as width, length, the height of the front evaporator, and roof transparency. A suitable structure was obtained for the multi-layer perceptron (MLP) method and the mathematical statistics % AARE, RMSE, and R2, were used to evaluate network performance. The method showed good agreement with the experimental data. Using the optimized created network, the effect of each parameter on the produced fresh water was assessed. Finally, a 125 m wide and 200 m long of greenhouse with a 4 m height of front evaporator and roof transparency of 0.6 that produced 161.6 m3/day of fresh water was introduced as the optimal seawater greenhouse.

Introduction

Fresh water scarcity in many countries, especially in North Africa and in the Middle East (MENA) has turned into a crisis. Many of these countries are faced with climate change, frequent drought occurrences and insufficient rainfall. A combination of various methods including water management and reduced water consumption, domestic and industrial wastewater treatment and desalination of salt and brackish water are used to cope with this problem. The saline and brakish water desalination using renewable energy is the first step that may somewhat compensate for the fresh water scarcity. In recent years, membrane technology, especially reverse osmosis units have been grown for the agriculture purposes. However, it must be noted that this method entails very high energy costs. The solar seawater greenhouse is one of the suitable and low-cost methods for the agriculture industry. The humidification-dehumidification process has been utilized in the solar seawater greenhouses. This method has some advantages including flexible capacity, low installation and operational costs, ease of use and utilization of clean and renewable energies (Campos et al., 2017). Numerous studies have been conducted on the solar seawater greenhouses. The most studies are thermodynamic modeling, while limited experimental studies have been carried out. In the most of the studies tried to investigate the effect of geometry dimensions on the energy consumption and water production in the solar seawater greenhouse. Goosen et al. (2003) conducted an empirical study to investigate the effects of various greenhouse parameters on the fresh water production in the desalination unit of a seawater greenhouse. They also used a thermodynamic model based on heat and mass balances. They noticed that the greenhouse dimensions greatly influenced water production and energy costs. Three different weather conditions were assumed for the interior of the greenhouse: normal climate, tropical climate and desert climate. Results showed that the largest volume of fresh water was produced at desert climate and the least amount of energy was consumed at tropical climate.

Davies and Paton (2004) studied the energy model of a seawater greenhouse in the United Arab Emirates. They investigated the effects of three different shading materials (semi opaque sheets, perforated sheets, and pipe arrays) on fresh water production and temperature drop in the greenhouse. Results indicated that the semi opaque sheets has slightly improved water production and cooling, whereas the perforated sheets significantly increased fresh water production. Pipe arrays considerably increased fresh water production and slightly decreased the temperature within the greenhouse.

One type of these systems was studied for use in warm and arid areas by (Perret et al., 2005). They used two evaporators to increase ambient air relative humidity in the greenhouse at Sultan Qaboos University in Oman and also employed two condensers for water condensation. Results showed that water humidity after passing through the second evaporator increased to the saturation point, and water temperature in the condenser was always lower than the dew point of the air flow. Therefore, the condensation process occurred in the condenser. The low rate of condensation in the condenser was due to the high speed of the inflow air that did not allow sufficient contact time between the air and the condenser. The condensers of the system were also designed on wheels so that they could be moved easily to different areas in the greenhouse. Tahri et al. (2009) introduced a mathematical model based on energy and mass equations for a seawater greenhouse condenser producing fresh water in Muscat (Oman). The condenser dimensions were 15 × 1.9 × 0.8 m and 320 rows of pipes with a 30° angle between the pipes and the direction of the air inflow. Each array had 14 identical vertical pipes with the diameter of 32 mm, thickness of 200 μm, and height of 1.8 m.

Using combined energies in the seawater greenhouse system is also an attractive issue. Mahmoudi et al. (2008) employed a wind-solar hybrid system for providing fresh water without using fossil fuels. Results indicated that it was possible to produce about 297 m3/day of fresh water in a 60 m long and 16 m wide greenhouse.

The performance of the seawater greenhouse depends strongly on the condenser performance. In humidification dehumidification process which is used in the seawater greenhouse, the lower temperature of cooling water in the condenser caused to produce more fresh water. Dawoud et al. (2006) studied possible strategies for cooling seawater condensers. These techniques are evaporative cooling for surface seawater, make use of a cooling machine to cool the condenser coolant in a closed loop or utilize deep seawater as a condenser coolant.

Tahri et al. (2013) presented two models for simulation of condensation process of humid air in a seawater greenhouse condenser. In fact, they intended to develop a mathematical model based on mass transfer. They reported the effects of relative humidity, air temperature, seawater temperature, air velocity, and radiation rate on the amounts of condensation. Al-Ismaili et al. (2018) was also proposed some empirical correlations for the condenser of seawater greenhouse.

Recently, a new empirical model based on the composite desirability function methodology has been proposed. Yetilmezsoy and Abdul-Wahab (2014) introduced an empirical model to predict the quantity of produced water. Their presented formula was able to predict the performance of the system well. Zarei et al. (2018) used support vector regression to study the performance of a seawater greenhouse system and the parameters that influenced it.

The brine management is another challenges in desalination units. Akinaga et al. (2018) proposed the use of evaporative coolers in seawater greenhouses to reduce the volume of brine because this allowed cultivation of high value crops and production of sea salt. Unlike typical greenhouses, only natural wind was used for ventilation instead of electric fans. They introduced a model to predict water evaporation, salt production, temperature and humidity in the greenhouse according to ambient conditions.

As mentioned above, The function of the seawater greenhouse is influenced by various parameters. The relations between the variables are complex. Therefore, the artificial neural network method can be a powerful tool to predict the seawater greenhouse performance.

The main advantages of using Artificial Neural Networks (ANN) include: it can handle large amount of data sets; it has the ability to implicitly detect complex nonlinear relationships between dependent and independent variables; it has ability to detect all possible interactions between predictor variables. Until now, there is no research on the use of ANN on the seawater greenhouse variables. Therefore, the present research tried to use the intelligent method of neural networks and the multi-layer perceptron method to simulate parameters that influenced seawater greenhouses, based on available data. The general goal of the study was to investigate the effects of changes in the length and width of greenhouse, height of evaporators and roof transparency on the amount of the produced fresh water. The special and practical purpose of this study was to determine the optimal conditions for producing the largest amount of water in a specific greenhouse.

Section snippets

Seawater greenhouse mechanism

A seawater greenhouse is a type of desalination unit that uses sunlight and seawater, humidifies the air in the interior of the greenhouse and produces the fresh water. This fresh water can be used for irrigating and also drinking. Since seawater greenhouses use solar energy and employ fewer mechanical parts, they greatly reduce costs of energy consumption and construction, repair and maintenance compared to other desalination units. Greenhouses are built to create suitable environmental

Methodology

Modelling and process simulations are necessary for determining the effect of each parameter on system performance. Thermodynamic modeling based on heat and mass balances needs to solve a set of many equations which is difficult and time-consuming. Therefore, intelligent methods were used in this study to reduce computation time and increase computational accuracy. Artificial neural networks are capable of modeling highly nonlinear systems that are black boxes.

Results and discussion

In the present study, the data presented by (Goosen et al., 2003) was used to simulate the parameters of the greenhouse desalination system. They intended to use a desalination system based on humidification-dehumidification in dry coastal areas plagued with saline soil and scarcity of drinkable groundwater. Their specific objective was to determine parameters related to seawater greenhouses that were intended to produce fresh water and also grow agricultural crops in a greenhouse system. They

Conclusions

Seawater greenhouses can be a suitable agricultural strategy for regions facing water stress because they allow farming in remote arid coastal areas that usually cannot be used for crop production. In these greenhouses, the humidification-dehumidification process is employed for desalination. It is an inexpensive process with low installation and maintenance costs. It is simple to operate and suitably used in seawater greenhouses. The present research used an artificial neural network to study

Cited by (0)

View full text