Energy Use Pattern and Optimization of Energy Consumption for Greenhouse Cucumber Production in Iran Using Data Envelopment Analysis ( DEA )

In this study a non-parametric method of Data Envelopment Analysis (DEA) is used to estimate the energy efficiencies of cucumber producers based on eight energy inputs including human power, diesel fuel, machinery, fertilizers, chemicals, water for irrigation, electricity and seed energy and single output of production yield. Data were collected using face-to-face surveys from 25 greenhouses in Khuzestan province of Iran. Energy indices, technical, pure technical and scale efficiencies were calculated by using Data Envelopment Analysis (DEA) approach for 25 cucumber greenhouses. Total energy input and output were calculated as 163994 MJha and 62496 MJha, respectively, whereas diesel fuel consumption with 45.15% was the highest level between energy inputs. Energy output-input ratio, energy productivity and net energy gain were 0.38, 0.47 kg MJ, -101498 MJ ha, respectively. The average values of TE, PTE and SE were 88%, 91% and 96%, respectively.


Introduction
The high rate of population growth and reducing the extent of fertile land due to the increasing development of urban and industrial areas induce more efficient use of existing facilities.The effective and efficient use of limited resources like water, soil and human power that are of particular importance to provide food requirements for people in developing countries, including Iran.Successful efforts to achieve self sufficiency and growth of gross national income like any other activity requiring deep knowledge of the practical and economic processes and applying the latest knowledge and technology around the world.Greenhouse production technology led to increase the efficiency of limited water and soil resources.And its importance is undeniable with respect to the dry climate and low rainfall in most parts of Iran.The major disadvantage of this method is high energy consumption because in most cases greenhouse production is off-season.Increase in energy efficiency in greenhouse cultures is of the most important energy studies in agriculture, and any success in increasing energy efficiency in greenhouse cultures can cause efficient use of valuable energy resources.
In recent years, Data Envelopment Analysis (DEA) has become a central technique in productivity and efficiency analysis applied in different aspects of economics and management science that helps us to manage efficient use of resources and ultimately more profit.The DEA is a non-parametric method for estimating the production function.The major drawback of these methods is initial necessary for the production function consequently parametric methods are not suiTable for evaluation the units under control that may be inconsistent with the nature of the units under evaluation (Gheisari et al., 2007).Also in recent years, many authors applied DEA in agricultural enterprises; such as: evaluation efficiency of greenhouse strawberry (Banaeian et al., 2011), optimization of energy consumption for apple production (Mousavi-Avval et al., 2011a ), a comparative study of parametric and non-parametric energy use efficiency in paddy production ( Nassiri and singh, 2010), energy use pattern and benchmarking of selected greenhouses in Iran (Omid et al., 2011), study on energy use pattern and efficiency of corn silage in Iran (Pishgar komleh et al., 2011), analysis farming system in citrus farming in Spain (Reig-Martinez and Picazo-Tadeo, 2004), improving energy use efficiency of canola production (Mousavi-Avval et al., 2011b), energy use efficiency for walnut producers (Banaeian et al., 2010).
The aim of this research was to determine energy use pattern and energy use efficiency in the cucumber greenhouses in Khuzestan province using data envelopment analysis (DEA) and presentation methods for optimization energy consumption.

Selection of case study region and data collection
This study was conducted in Khuzestan province of Iran.This province is located within 29°58′ and 30°04′ North latitude and 47°41′ and 50°39′ east longitude, in the south-west of Iran.In this province, there are 38 vegeTable greenhouses (Anonymous, 2010).In this study, 25 active vegeTable greenhouses were studied to determine energy use and to evaluate the performance for greenhouse cucumber production.
Data were collected through personal interview method in a specially designed schedule for this study.The collected data belonged to the 2010/11 production year.

Energy equivalents of input and output
The data included the quantity of various energy inputs used per hectare of greenhouse cucumber production including: human power, machinery, diesel fuel, chemicals, water for irrigation, electricity, fertilizers and seed, and the production yield as output.In order to analysis the performance of greenhouses from an energy use efficiency point of view, all of inputs and output were then converted into energy equivalents by multiplying the quantity of input use with their corresponding energy equivalent coefficients.Energy equivalents, shown in Table 1, were used for estimation; these coefficients were adapted from several literature sources that best fit the conditions in Iran.
Following the calculation of energy input and output equivalents, the indices of energy consumption including energy ratio, energy productivity and net energy gain were estimated using the following Eqs.(Rafiee et al., 2010): (1) (2) (3)

Data envelopment analysis technique
Data envelopment analysis (DEA) is an analysis method to measure the relative efficiency a homogeneous number of organizations that essentially perform the same tasks (Cooper et al., 2006).In this case, they are cucumber greenhouses.So, the values of energy consumed from different energy inputs (MJ ha −1 ), as mentioned above, were defined as input parameters, and the yield of greenhouse cucumber production (kg ha −1 ) was defined as output parameter; also, each greenhouse was called a decision making unit (DMU).
In DEA, an inefficient DMU can be made efficient either by reducing the input levels while holding the outputs constant (input oriented); or symmetrically, by increasing the output levels while holding the inputs constant (output oriented) (Mousavi-Avval et al., 2011b).The choice between input and output orientation depends on the unique characteristics of the set of DMUs under study.In this study the input oriented approach was deemed to be more appropriate because there is only one output while multiple inputs are used; also as a recommendation, input conservation for given outputs seems to be a more reasonable logic (Galanopoulos et al., 2006); so the greenhouse cucumber yield is hold fixed and the quantity of inputs energy were reduced.
DEA has two models including CCR and BCC models.The CCR model is built on the assumption of constant returns to scale (CRS) of activities, but the BCC model is built on the assumption of variable returns to scale (VRS) of activities.Efficiency by DEA is defined in three different forms: overall technical efficiency ( CCR TE ), pure technical efficiency ( BCC TE ) and scale efficiency (SE).

Technical efficiency
Technical efficiency (TE) can be calculated by the ratio of sum of weighted outputs to sum of weighted inputs (Cooper et al., 2006): (4) Or mathematically as (Nassiri and Singh, 2009): (5) Where 'x' and 'y' are inputs and outputs, 'v' and 'u' are input and output weights, respectively, 's' is the number of inputs (s = 1, 2, …, m); 'r' is the number of outputs (r = 1, 2, ..., n); and 'j' represents j th DMU (j = 1, 2, ..., k).For solving Eq. ( 5 In other words decomposition of Eq. ( 15) can be defined by: (16) This decomposition, which is unique, depicts the sources of inefficiency, i.e., whether it is caused by inefficient operation (PTE) or by disadvantageous conditions displayed by the scale efficiency (SE) or by both.If the scale efficiency is less than 1, the DMU will be operating either at decreasing returns to scale (DRS) if a proportional increase of all input levels produces a less-than-proportional increase in output levels or increasing return to scale (IRS) at the converse case.This implies that resources may be transferred from DMUs operating at DRS to scale to those operating at IRS to increase average productivity at both sets of DMUs (Pishgar komleh et al., 2011).By solving of CCR and BCC models, the weights of remaining inputs (diesel fuel, chemicals, electricity, machinery, water for irrigation, human power, fertilizers and seed) and output (greenhouse cucumber) would be calculated so the maximum value of is calculated.
The data analysis was carried out with help of the Excel 2007 spreadsheet and Frontier Analyst Professional software.

Analysis of energy input and output in greenhouse cucumber production
Amount of inputs, output and their energy equivalents for greenhouse cucumber production is presented in Table 2.The total energy consumption for greenhouse cucumber production was calculated as 163994 MJ ha −1 ; also, the percentage distribution of the energy associated with the inputs is seen in Table 3.It is evident that, the greatest part of total energy input (45.15%) was consumed by diesel fuel consumption.Also, fertilizers and seed were the second main energy consuming input.Similar studies had also reported that diesel fuel and fertilizers were the most intensive energy inputs (Zangeneh et al., 2010;Esengun et al., 2007;Cetin and Vardar, 2008;Ghasemi Mobtaker et al., 2010).ListenRead phoneticallyIn Khuzestan province is used -however short term-of heating systems in greenhouses due to the large temperature difference between day and night and the low temperature at night.Therefore diesel fuel consumption is allocated to the share largest from other inputs.In order to improve the greenhouse environment as well as reduction of diesel fuel consumption, it is strongly suggested that the heating system efficiency is raised or replaced with alternative sources of energy such as natural gas and solar energy (Omid et al., 2011).
The results also revealed that electricity was the third main energy consuming input because of rising temperatures on some days; the ventilation system is used to regulate the greenhouse temperature.ListenRead phonet the water for irrigation was the least energy demanding inputs for greenhouse cucumber production.On the other hand, the average cucumber yield obtained was found to be 78120 kg ha −1 ; accordingly, the total energy output was calculated as 62496 MJ ha −1 , in the enterprises that were analyzed.In the previous study on greenhouse cucumber production in Tehran province of Iran the yield value of greenhouse cucumber and total output energy were reported higher than that of this study (Omid et al., 2011).The lower yield value and energy output of greenhouse cucumber production in Khuzestan province were mainly due to the mismanagement of input usage.
The energy output-input ratio, energy productivity and net energy gain of greenhouse cucumber production are presented in Table 3. Energy ratio was calculated as 0.38, showing the inefficiency use of energy in greenhouse cucumber production in Khuzestan province.It is concluded that the energy ratio can be increased by raising the crop yield and/or by decreasing energy input consumption.Similar results obtain 0.64 for the energy ratio of greenhouse cucumber production (Omid et al., 2011;Mohammadi and Omid, 2010).The average energy productivity of greenhouse cucumber production was 0.47 kg MJ −1 .This means that 0.47 units output was obtained per unit energy.Similar results have been reported 0.39 and 0.8 kg MJ −1 for the energy productivity of greenhouse cucumber production (Mohammadi and Omid, 2010).The net energy gain of greenhouse cucumber production was 101498 MJ ha −1 .Net energy gain is negative (less than zero).Therefore, it can be concluded that in greenhouse cucumber production, energy is being lost.Similar results obtain 53027.16MJ ha −1 and 55552.83MJ ha −1 for the net energy gain of greenhouse cucumber production (Mohammadi and Omid, 2010;Omid et al., 2011).
The distribution of inputs used for greenhouse cucumber production in groups of direct, indirect, renewable, and non-renewable sources is shown in Table 3.The ratio of direct and indirect energy sources are 67.52% and 32.47%, respectively.Also, there is a significant difference between renewable and non-renewable energy sources.Renewable energy sources are clean sources of energy that have a much lower impact on the environment than do conventional energy technologies.In the studied greenhouses, 94.98% of the input energy comes from non-renewable energy sources, which are finite and will someday be depleted.Also, many of these energy sources are harmful to the environment (Unakitan et al., 2010).Several researchers showed that the ratio of direct energy is higher than that of indirect energy, and the rate of non-renewable was much greater than that of renewable consumption in cropping systems (mohammadi et al., 2008;Hatirli et al., 2006).

Energy use efficiency for unit greenhouses
In this study, we used CCR and BCC models to evaluate technical, pure technical and scale efficiencies (TE, PTE and SE, respectively) of cucumber greenhouses.The results of CCR and BCC models are shown in Table 4. Based on CCR results, this study shows that only 4 greenhouses were relatively efficient and the remaining 21 where inefficient, i.e. their efficiency score were below 1.But from the results of BCC model 6 greenhouses (out of total 25 greenhouses) were efficient, meaning they have an efficiency score of 1(Table 4).Other greenhouses who have efficiency score less than 1, are inefficient in energy use.The average values of the technical efficiency, pure technical efficiency and scale efficiency are summarized in Table 5.The average values (for all 25 greenhouses considered) of TE, PTE and SE were found to be 88%, 91% and 96%, respectively.The average TE of the inefficient DMUs at the greenhouse unit was calculated as 88%.This implies that the same level of output could be produced with 88% of the resources if these units were performing on the frontier.Another interpretation of this result is that 12% of overall resources could be saved by raising the performance of these DMUs to the highest level.This indicates that there is ample scope in improving inefficient greenhouses for operating practices to enhance their energy use efficiency.In a similar study, TE, PTE and SE for cucumber greenhouse were reported to be 87%, 97% and 90%, respectively (Omid et al., 2011).

Reference units with coefficients of decision
The efficient greenhouses obviously follow good operating practices.However, among the efficient greenhouses, some (greenhouses: 6, 11, 12 and 21) show better operating practices than others.Therefore, discrimination is required to be made among the efficient greenhouses while seeking the best operating practices.These efficient greenhouses can be selected by inefficient DMUs as best practice DMUs.
So for each inefficient unit an efficient unit or combination of efficient multi-unit as reference units is introduced.Each share of efficient units in the reference pattern for an inefficient unit depends on the weight of which is calculated for each of the efficient units using DEA approach (Emami meibodi, 2000).
For example, in the case of the greenhouse 5, the reference composite DMU is formed by the greenhouse 6 (Table 4, CCR model).This means the greenhouse 5 is close to the efficient frontier segment formed by this efficient DMU.The production efficiency can be obtained for greenhouse 5 with the introduction of efficient greenhouse reference 6.The selection of this efficient DMU is made on the basis of its comparable level of inputs and output yield to the greenhouse 5.In Table 4 (CCR model), the reference unit for greenhouse 5 is expressed as 6(81.7%),where 6 is the DMU number while the value between brackets is weight of reference greenhouse that shows the contribution amount of reference greenhouse 6 in evaluation of inefficient greenhouse 5.The 91.69% efficiency of the greenhouse 5 means that for being an efficient unit, it has to decrease 8.31% of all inputs (without decreasing output).On the other hand, needed inputs for a specific level of output can be calculated; by considering the greenhouse 6 as the pattern of the greenhouse 5, and decision coefficient of the greenhouse 5 in the Table 4 which is 81.7%.So to have the greenhouse 5 efficient, it has to use 81.7% of the inputs of unit 6, without decreasing output.

Returns to scale
The latest column of Table 5 indicated results of return to scale.The analysis shows that DMUs numbered 6, 11, 12 and 21 that are efficient under the CRS model are both technically and scale efficient (Table 5).
The RTS indicated that all efficient units (based on technical efficiency) were operating at CRS, whereas all inefficient ones were at IRS, which indicates that for considerable changes in yield, technological change is required.

Slack and surplus energy consumption in each of greenhouses
According to the results obtained from Table 6 greenhouses 6, 11, 12 and 21 have constant efficiency but the remains have increasing efficiency.Table 6 shows the obtained results from analyzing greenhouses by using input oriented constant returns to scale model.Data of this Table are used for determining extra input and deficiency of efficiency.The specific quantity of input that each inefficient unit needs to decrease in order to become efficient is determined.As Table 6 shows, the greenhouse 1 with the efficiency of 91.47% has to decrease 22154 units of diesel fuel, 9540 units of electricity and 7894 units of fertilizers and seed to stand on the efficiency partition line.The average share of each input in decreasing energy consumption for greenhouses is illustrated in the Figure 1.

Conclusion
This study described an in-depth application of input oriented DEA model to investigate the degree of technical and scale efficiency of 25 cucumber greenhouses in Khuzestan province of Iran.This procedure allows the determination of greenhouses 6, 11, 12, and 21 as the best practice greenhouses that can be providing useful insights for other greenhouse management.Diesel fuel, total fertilizers and electricity energy inputs had the highest potential for saving energy; so, if inefficient greenhouses would pay more attention towards these sources, they would considerably improve their energy productivity.
) the following linear program (LP) was developed by Charnes et al., which called CCR model: technical efficiency ( CCR TE ) and i represents DMU (Houshyar et al., 2010).2.3.2Pure technical efficiency In 1984, Banker et al. introduced a model in DEA, which was called BCC model to draw out the technical efficiency of DMUs.The calculation of efficiency in BBC model is called Pure Technical Efficiency and can be expressed by Dual Linear Program (DLP) as(Houshyar et al., 2010): CCR and BCC scores, scale efficiency defined by(Cooper et al., 2006):   (15)

Table 1 .
Energy equivalent of energy output and input in agricultural production

Table 2 .
Energy used status for cucumber production in Khuzestan province

Table 3 .
Energy output-input ratio and forms in greenhouse cucumber production

Table 4 .
Evaluation of cucumber greenhouses with reference units via CCR and BCC input oriented models

Table 5 .
Analysis of efficiency and return to scale in greenhouse cucumber production in Khuzestan province

Table 6 .
Slack and surplus energy consumption in each of greenhouses with CCR input oriented model (MJha -1 ) Figure 1.Energy saved via each input with CCR input oriented in cucumber greenhouses