Improving Energy Efficiency and Greenhouse Gas Emissions in Small Farm Wheat Production Scenarios Using Data Envelopment Analysis

Abstract

Assessing the energy cycle and greenhouse gas (GHG) emissions of wheat production in small Egyptian farms is essential to improve wheat productivity to meet population growth and achieve sustainable development. This study aims to compare wheat production in terms of energy use and GHG emissions for different scenarios in the Delta of Egypt and to use Data Envelopment Analysis (DEA) to optimize the wheat production system. Three common scenarios of the wheat production system (S-I, S-II, and S-III) from old lands with one scenario (S-IV) from newly reclaimed land were included in the study. Data were collected from small farmers through a face-to-face questionnaire and interviews in 2022–2023. The results showed that the third scenario (S-III) in the old lands had the lowest input energy consumption (42,555 MJ ha⁻¹) and the highest output energy (160,418 MJ ha⁻¹), with an energy use efficiency of 3.770. In comparison, the input and output energy for the newly reclaimed scenario (S-IV) were 37,575 and 130,581 MJ ha⁻¹, respectively, with an energy use efficiency of 3.475. S-III was an optimum scenario due to its high energy indicators, such as energy productivity of 0.173 kg MJ⁻¹. The total GHG emissions of S-III were the lowest in old lands with a value of 1432.9 kg CO_2-eq ha⁻¹, while S-IV had 1290.2 kg CO_2-eq ha⁻¹. The highest GHG emissions input was diesel fuel for machinery and irrigation, followed by manure, chemical fertilizers, and agricultural machinery use. Using mechanization in most farming operations for S-III and S-IV led to decreased losses of agricultural inputs with increasing outputs (yield and straw). Therefore, using them in wheat farming practices is recommended to increase the wheat farming system’s energy efficiency and GHG emissions.

1. Introduction

Wheat is the most critical food grain crop on which the Egyptian people depend for food, e.g., producing bread and pasta from grain and using wheat straw as a primary animal feed. The average wheat production in Egypt is about 9.7 Gg in 2022, which is higher than the previous year by 7% and nearly higher than the five-year average by 10% [1]. The smallholding farms (<4 ha) dominate 63% of the total cultivated area, and 64.1% of the wheat production area is located in the Egyptian Delta. However, smallholder farmers suffer from a labor shortage, climate change, and the lack of appropriate agricultural machinery for their farming operations [2]. In order to fulfill the demand for wheat and its products and the continual population growth, Egypt places a high value on the wheat crop and actively encourages farmers to cultivate it. As a result, to cut GHG emissions, wheat growers must use energy effectively.

Energy is a necessary factor in agriculture, and the significant population increase leads to a rise in water and food consumption and an increase in energy use, which leads to severe economic and environmental hardships [3]. The agricultural sector’s energy consumption is increasing globally [4]. The agricultural sector provides various products and services and is critical to human survival. Since the 20th century, the imbalance between the low land productivity and the high-density population has severely hampered agricultural development. Large and intensive farming was implemented using highly chemical fertilizers and pesticides to increase productivity and address the gap between food and population growth [5]. This led to negative environmental consequences such as natural resource damage, decreased land productivity, an increase in pests and diseases on the land, reduced land biodiversity, and increased energy use [6]. Increasing energy use efficiency reduces GHG emissions and sustains food security. Several studies that evaluated the productivity of various entities employing Data Envelopment Analysis (DEA) in agricultural production demonstrated crucial advancements that help create an efficient production system [7,8,9,10].

Around 10–12% of GHG emissions worldwide are caused by agricultural operations, which grabs the attention of global policymakers and planners [11]. Climate change has become a concern in Egypt, as Egypt’s carbon production in 2019 amounted to about 246.64 Gg of carbon, which represents 0.68% of global output at 2.46 Gg of carbon per person [12]. Therefore, the Egyptian government is working to inventory data and sources of GHG emissions to design sustainable agricultural systems that reduce vulnerability to climate change by strengthening agricultural policy, as the farming sector is one of the largest producers of GHG emissions. Therefore, analyzing the energy use behavior and GHG emissions in crop production is essential for a complete assessment of the agro-environmental structure [13].

Crop production is strongly related to different energy sources, such as labor, fuel, and machinery. Also, agricultural inputs produce GHG emissions during agricultural operations [14]. So, identifying energy consumption and efficiency is a critical process to assess input and output energy in crop production [7]. Also, efficient energy use is fundamental for decreasing environmental loss, protecting resources, extending agricultural sustainability development, and reducing GHG emissions [15]. As a result, a reasonable action toward reducing GHG emissions is to reduce the emissions from production processes and identify the most economical options that achieve sustainability for GHG emissions reduction [16]. Several studies have assessed wheat production’s energy use, efficiency, and GHG emissions [17,18]. However, there are not enough studies to quantify the GHG emissions and energy consumption of wheat in Egypt. Therefore, this work aims to (i) assess total energy use for different wheat farming scenarios in small farms, (ii) calculate the GHG emissions caused by energy use, and (iii) use DEA to improve energy use efficiency and reduce GHG emissions.

2. Materials and Methods

2.1. Study Area

The study was carried out in the Egyptian Delta in 2022–2023, which is in Egypt’s northern region, as shown in Figure 1. The Egyptian Delta lies between longitudes from 29°37′54″ E to 32°55′49″ E and latitudes from 29°37′49″ N to 31°41′2″ N. In the Delta of Egypt, the average arable land area and the number of householders in small areas (less than 2 ha) are 964,550 hectares and 2,643,124 households, respectively. The Delta region has a mild climate with an average temperature of 25 degrees; from 100 to 200 mm of rain falls during an average year, and most of this falls in the winter months. Winter wheat production dominates most of its arable lands, so wheat was selected for the study.

2.2. Wheat Production System in the Delta of Egypt

Organic fertilizers and superphosphate are added before tillage, and soil tillage operations are carried out during September. Wheat is sown in October, and nitrate, potassium, and phosphorous fertilizers are also applied. Chemical control by spraying is used against diseases and pests twice to thrice during the growing season before wheat harvesting in June and July. Typically, the crop is irrigated by immersion six to ten times during growth.

Concerning agricultural machinery use, the arable lands of the Egyptian Delta are divided into old lands (clay lands) and newly reclaimed lands (sandy lands). The scenarios used in the performance of the agricultural operations of wheat in the small farms in the study area were studied in 11 scenarios (for more details, see Sayed et al. [19]). In this study, through a randomly selected questionnaire, data were obtained from farmers and divided into 4 different scenarios according to the method of performing agricultural operations (whether using mechanization or labor). These 4 scenarios are most common in the Delta of Egypt, including 3 scenarios in the old lands and one in the newly reclaimed lands, as shown in

Table 1. Scenarios of wheat agricultural operations in Delta small farms.

Scenario * (Small Farm)	Wheat Agricultural Operations on Old Lands
	Organic Fertilization	Tillage	Sowing	Chemical Fertilization	Spraying	Harvesting	Threshing
S-I (<1 ha)	◊	√	◊	◊	√	◊	√
S-II (1–1.5 ha)	◊	√	◊	◊	√	√	√
S-III (1.5–2 ha)	√	√	◊	◊	√	√	√
	Wheat Agricultural Operations on Newly Reclaimed Lands
S-IV (<2 ha)	√	√	√	◊	√	√	√

* Using labor (◊) and using mechanization (√).

.

The small wheat farmers were surveyed face-to-face to get data on the wheat farming system. The Cochrane equation and a straightforward random sampling technique were used to determine the suitable sample size. The required sample size was calculated by Equation (1) [20]:

$$n=\frac{N{(Z\times \sigma )}^2}{{(N-1)e^2+Z\sigma }^2}$$

(1)

where $n$ is the total sample size of data; $N$ is the whole population number; $\sigma$ is the standard deviation of the sample; $Z$ is the coefficient of reliability (1.96 represents 95% reliability); and $e$ is the acceptable sampling error (5%).

Accordingly, 385 farmers were selected, including 131 farmers from S-I, 117 from S-II, 94 from S-III, and 43 from S-IV. All 385 farms in the Egyptian Delta have the same geographical and climate characteristics but different farming systems. S-I is the most prevalent in the Egyptian Delta because most of its areas are less than a hectare (<1 ha), and this leads to the non-use of agricultural machinery in some farming operations.

2.3. Energy Use Analysis

This study evaluated small wheat farms (<2 ha). The data included information on inputs (agricultural machinery, human labor, chemical fertilizers, seed, diesel fuel, etc.) and outputs (grain and straw yield). The energy requirements and input quantity for each item were determined at each major stage of wheat production, from land preparation to wheat harvesting. During wheat growth, common inputs include fertilizers, seeds, labor, chemical poison, diesel fuel, machinery, and water. At the same time, the outputs are the weight of the resulting wheat crop, which includes yield and straw. Regarding the system boundaries, the study evaluated the wheat production system starting from preparing the soil for planting to wheat harvest, as shown in Figure 2.

The inputs and outputs of the study were converted to MJ ha⁻¹ using the energy equivalents. The amount of all energy inputs, such as human labor, fertilizers, diesel fuel, chemical poison, water, seeds, etc., was calculated by multiplying energy equivalents in

Table 2. Energy equivalents of the wheat production system’s inputs and outputs.

Item	Unit	Energy Equivalent (MJ unit−1)	Reference
Inputs
1. Human labor	h	1.96	[21,22,23]
2. Agricultural machinery	kg	138	[24]
3. Diesel fuel	L	56.31	[23,25,26]
4. Chemical fertilizers
(a) Nitrogen (N)	kg	66.14	[22,27,28]
(b) Phosphorus (P)		12.44
(c) Potassium (K)		11.15
5. Manure	kg	0.3	[21,23,29]
6. Chemical poison
(a) Herbicide	L	238	[26,30,31]
(b) Pesticide		199	[30]
(c) Fungicide		92	[30]
7. Water for Irrigation	m3	1.02	[32]
8. Seeds	kg	15.7	[26]
Outputs
1. Wheat grain yield	kg	14.7	[26,33]
2. Wheat straw yield	kg	9.25	[34]

with the energy input items. The total energy in the study was calculated in the following categories:

2.3.1. Fuel Consumption (FC)

Fuel is used inside the farm in irrigation operations and in tractors and agricultural machinery used to perform agricultural operations. In this study, based on the application of information from expert farmers, the consumption of diesel fuel used for irrigation (FC_irrigation) was calculated using the following equation:

$${FC}_{irrigation}=N\times F_O\times E_f$$

(2)

As for the diesel fuel consumption in tractors and agricultural machinery (FC_tractor), it was calculated using the following equation:

$${FC}_{tractor}=F_{hr}\times t\times E_f$$

(3)

where N is the number of irrigation times required for wheat during the season per hectare; F_o is the fuel required per one irrigation (liters per one irrigation); F_hr is the fuel required to operate the tractor within one hour (liters per hour); t is the tractor operating time (hours per hectare); and E_f is the equivalent of energy (MJ L⁻¹).

2.3.2. Energy for Machinery

According to Mousavi-Avval et al. [8], the indirect energy used in the manufacture of each machine in wheat production is calculated from the following equation:

$${ M}_E=E\times \frac{W}{T}\times Q_h$$

(4)

where M_E is the energy of machinery (MJ ha⁻¹); E is the energy that a machine provided (MJ kg⁻¹); W is the specific machine weight (kg); Q_h is the total working hours for the machine during the agricultural season (h ha⁻¹); and T is the machinery lifetime (h). The ASAE standards and the local use pattern were considered when determining the average lifetime of 12,000 h [35].

2.3.3. Human Labor

The capacity to exert an average amount of energy throughout one hour of agricultural activity is known as muscular strength. Human labor equivalent energy is the muscular force used in agricultural operations to produce the crop. To accomplish this, the number of workers required for each agricultural operation and the duration of a worker’s shift were investigated. As a result, human work energy was computed using the relevant energy equivalent.

2.3.4. Spraying

To calculate the pesticide energy in agricultural production, the amount of active ingredient (L ha⁻¹) is multiplied by the equivalent energy used to produce it.

2.3.5. Fertilizers

Based on the N, P, and K energy equivalents, the calculation of the fertilizer’s energy was as follows:

$${ E}_f=W_f\times E_k$$

(5)

Where E_f is the energy of fertilizers (MJ ha⁻¹); W_f is the weight of consumed fertilizers (kg ha⁻¹); and E_k is the energy of fertilizers per kg (MJ kg⁻¹).

Evaluating the agricultural system’s energy efficiencies is critical using the energy ratio of outputs and inputs. The main four indices of energy were used to analyze the energy efficiency of agricultural inputs [33]:

$$Energy use Efficiency=\frac{Energy output (\mathrm{M}\mathrm{J} {\mathrm{h}\mathrm{a}}^{-1})}{Energy input (\mathrm{M}\mathrm{J} {\mathrm{h}\mathrm{a}}^{-1})}$$

(6)

$$Energy Productivity=\frac{Crop yield (\mathrm{k}\mathrm{g} {\mathrm{h}\mathrm{a}}^{-1})}{Energy input (\mathrm{M}\mathrm{J} {\mathrm{h}\mathrm{a}}^{-1})}$$

(7)

$$Specific Energy=\frac{Energy input (\mathrm{M}\mathrm{J} {\mathrm{h}\mathrm{a}}^{-1})}{Crop yield (\mathrm{k}\mathrm{g} {\mathrm{h}\mathrm{a}}^{-1})}$$

(8)

$$Net Energy=Energy output \left(\mathrm{M}\mathrm{J} {\mathrm{h}\mathrm{a}}^{-1}\right)-Energy input \left(\mathrm{M}\mathrm{J} {\mathrm{h}\mathrm{a}}^{-1}\right)$$

(9)

2.4. Data Envelopment Analysis (DEA)

DEA is a data-oriented technique used to estimate resource use efficiency and rank production units based on their performances. Charnes–Cooper–Rhodes (CCR) developed a common arithmetic approach called DEA, which was continued by Banker–Charnes–Cooper (BCC) [36,37]. Furthermore, economists are increasingly using DEA to evaluate energy usage effectiveness and the environmental effects of various production methods [38]. Technical efficiency (TE) can be evaluated using DEA. Moreover, Decision Making Units (DMUs) can be assessed using DEA, and efficient DMUs could generate more outcomes with the same number of inputs as other DMUs. Manure, seeds, chemical fertilizers, pesticides, diesel fuel for the pump, irrigation water, and labor were employed as inputs in this study, while the outputs were grain wheat and straw. Data envelopment analysis was done using the Frontier Analyst version 4 software. The DEA was used to examine the energy use trend in the old Delta lands. The weighted outputs sum was divided by the weighted inputs sum to determine the TE as follows:

$${TE}_j=\frac{{\alpha }_1{x}_{1j}+{\alpha }_2{x}_{2j}+\cdots +{\alpha }_n{x}_{nj}}{{\beta }_1{y}_{1j}+{\beta }_2{y}_{2j}+\cdots +{\beta }_n{y}_{nj}}=\frac{{\sum }_{s=1}^m{\alpha }_1{x}_{1j}}{{\sum }_{r=1}^n{\beta }_1{y}_{1j}}$$

(10)

where TE_j is the technical efficiency rating of the jth DMU (ranging in value from 0 to 1); x and y correspond to output and input; $\alpha$ is the weights of output; $\beta$ is the weights of input; s is the output number (s = 1, 2,… m); r is the input number (r = 1, 2,… n); and j is the jth DMU (j = 1, 2, k).

The DEA model was applied to differentiate between inefficient and efficient DMUs, as well as to rate the inefficient DMUs. As a result, the energy saving target ratio (ESTR) helped to clarify the level of energy use inefficiency for the DMUs during the evaluation of efficient and inefficient DMUs. The jth DMU is represented by j in this calculation, and the energy-saving emphasis that might be achieved even without the request of decreasing output level is the overall drop in energy input variables. The minimum value of the energy saving target is zero; therefore, the values of ESTR_j will be between zero and one. The maximum values of ESTR indicate higher energy use inefficiency and, as a result, higher energy savings. The formula is given below:

$${ESTR}_j=\frac{Energy saviny target j}{Actual energy input j}$$

(11)

2.5. Estimation of GHG Emissions

In this study, GHG equivalents were used to calculate GHG emissions of production inputs (kg CO_2-eq ha⁻¹) such as diesel fuel, fertilizers, manure, pesticides, and machinery.

Table 3. The equivalents of CO₂ for wheat production inputs.

Input-Output (MJ)	Equivalent (kg CO2-eq MJ−1)	References
1. Machinery	0.071	[39]
2. Diesel fuel	2.76	[40]
3. Chemical fertilizers
(a) Nitrogen (N)	1.3	[41,42]
(b) Phosphorus (P)	0.2	[41,42]
(c) Potassium (K)	0.15	[41,42]
4. Manure	0.126	[43]
5. Chemical poison
(a) Herbicide	6.3	[41,42]
(b) Pesticides	5.1	[41,42]
(c) Fungicide	3.9	[41,42]

presents the GHG equivalents (kg CO_2-eq ha⁻¹). Furthermore, the GHG emissions per MJ were calculated to compare crop production reaction to emissions.

3. Results and Discussion

3.1. The Analysis of the Wheat Production System Energy

In the old lands of the Egyptian Delta, the three scenarios S-I, S-II, and S-III for agricultural wheat production were different in the total inputs and outputs of energy (

Table 4. Input and output energy consumption (MJ ha⁻¹) for the wheat production system.

Item	Old Lands						Newly Reclaimed Lands		Mean	SE	Share (%)
	S-I (<1 ha)		S-II (1–1.5 ha)		S-III (1.5–2 ha)		S-IV (<2 ha)
	Energy	%	Energy	%	Energy	%	Energy	%
Direct Inputs
1. Human labor	1051	2.29	687	1.54	498	1.17	219	0.58	614	349	1.4
2. Diesel fuel for machinery	6533	14.24	7371	16.48	8014	18.83	6681	17.78	7150	682	16.8
3. Diesel fuel for Irrigation	7036	15.33	6814	15.23	6250	14.69	5629	14.98	6432	629	15.1
4. Water	7098	15.47	6858	15.33	6579	15.46	6229	16.58	6691	374	15.7
Indirect Inputs
1. Agricultural machinery	1092	2.38	1351	3.02	1397	3.28	1289	3.43	1282	134	3.0
2. Chemical fertilizers
(a) Nitrogen (N)	15,741	34.31	14,639	32.73	13,065	30.70	11,806	31.42	13,813	1731	32.3
(b) Phosphorus (P)	740	1.61	711	1.59	651	1.53	592	1.58	674	66	1.6
(c) Potassium (K)	199	0.43	186	0.42	159	0.37	133	0.35	169	29	0.4
3. Manure	871	1.90	835	1.87	771	1.81	750	2.00	807	56	1.9
4. Seed	3827	8.34	3684	8.24	3636	8.54	2870	7.64	3504	430	8.2
5. Chemical poison
(a) Herbicide	906	1.98	850	1.90	821	1.93	736	1.96	828	71	1.9
(b) Pesticides	710	1.55	663	1.48	639	1.50	568	1.51	645	59	1.5
(c) Fungicide	79	0.17	79	0.18	74	0.17	72	0.19	76	3	0.2
Total Inputs	45,885		44,726		42,555		37,575		42,685	3676
Outputs
1. Wheat grain yield	89,949	64.49	96,212	65.00	108,457	67.61	85,891	65.78	95,127	9848	65.7
2. Wheat straw yield	49,534	35.51	51,812	35.00	51,961	32.39	44,690	34.22	49,499	3393	34.3
Total outputs	139,483		148,024		160,418		130,581		144,626	12,710

). As shown in Table 4, S-I had the highest input energy (45,885 MJ ha⁻¹) because it had the lowest use of agricultural machinery among the scenarios. Performing agricultural operations by labor increased the consumption of inputs such as chemical fertilizers, water, and seeds. At the same time, S-I had the lowest energy output (139,483 MJ ha⁻¹) because the low degree of mechanization decreased the productivity of grain and straw. In contrast, S-III had the lowest total energy input (42,555 MJ ha⁻¹) and the largest output energy (160,418 MJ ha⁻¹). This is due to depending on mechanization used in various agricultural operations, which decreases input consumption and increases grain and straw productivity.

In S-III, the highest input energy source among all others was diesel fuel for machinery and irrigation (33.52%), followed by chemical fertilizers (32.61%), water (15.46%), seeds (8.54%), chemical poison (3.61%), agricultural machinery (3.28%), manure (1.81%), and human labor (1.17%). The last arrangement of input energy items was the same as that obtained by Mondani et al. [16]. The difference between S-I, S-II, and S-III was due to the level of mechanization in each scenario.

Regarding direct energy inputs, human labor had the lowest value (498 MJ ha⁻¹) in S-III because most farm operations depended on mechanization. In comparison, the highest human labor energy was 1051 MJ ha⁻¹ in S-I due to mechanization only in three agricultural operations (tillage, spraying, and threshing). Considering diesel fuel for agricultural machinery, S-I had the lowest consumption (6533 MJ ha⁻¹) compared with the other scenarios. Of all scenarios, S-III had the largest diesel fuel energy consumption for machinery (8014 MJ ha⁻¹) due to the mechanization of most agricultural operations. In this regard, the total diesel fuel energy represents about 29.57, 31.71, and 33.52% of the total input energy for S-I, S-II, and S-III, respectively. It is also noted that the water input energy was the highest in S-I (7098 MJ ha⁻¹) because the workers’ poor distribution of organic fertilizers affected the water distribution movement inside the land. Therefore, the irrigation pump took more time during irrigation.

As for indirect energy inputs, chemical fertilizers ranked first for input energy in wheat production at 36.35 and 34.74% in S-I and S-II, respectively, and ranked second at 32.61% in S-III. The agricultural machinery differed in each scenario because of the difference in the number of mechanization operations. The agricultural machinery energy was 1397 MJ ha⁻¹ in S-III as the highest scenario, while it was the lowest in S-I at 1092 MJ ha⁻¹. The manure energy value differed between the scenarios—871, 835, and 771 MJ ha⁻¹ for S-I, S-II, and S-III, respectively—due to the difference between using labor and mechanizing the operation. As for the output energy, it was the highest in S-III (160,418 MJ ha⁻¹), while it was the lowest in S-I (139,483 MJ ha⁻¹). This is due to farmers’ use of harvesting machines in S-III, which reduces grain and straw losses, as mentioned in [44,45].

In the newly reclaimed lands, there is a more prevalent scenario (S-IV) in which farmers rely on mechanization of all agricultural operations except chemical fertilization. The total input and output energies were 37,575 and 130,581 MJ ha⁻¹, respectively. The input energy for chemical fertilization was highest among the other inputs in this scenario at 33.35%, followed by fuel for machinery and irrigation pumps (32.76%), water (16.58%), seeds (7.64%), and agricultural machinery (3.43%). The use of mechanization in S-IV reduced the amount of water, chemical fertilizers, and seeds required compared with the scenarios of the old lands due to the uniform distribution and reduced wastage of chemical fertilizers [46,47,48,49,50]. On the other hand, S-IV had lower output energy (130,581 MJ ha⁻¹) compared with the scenarios for old lands because the newly reclaimed lands were less fertile than the old lands, resulting in less grain and straw production.

3.2. The Agricultural Mechanization and Fuel Consumption Energy in the Wheat Production System

As shown in Figure 3a, the fuel energy was the highest in S-III at 14,265 MJ ha⁻¹, followed by S-II at 14,184 MJ ha⁻¹ and then S-I at 13,569 MJ ha⁻¹. This is because S-III depends on using agricultural mechanization in most agricultural operations. Farmers in the old lands rely on flood irrigation in the wheat irrigation method, while in the newly reclaimed lands, most farmers depend on modern irrigation. Among the scenarios, irrigation was the highest energy consuming, as it constituted 48% and 46% of the average total operating energy as fuel in the old and new lands, respectively, which is in line with earlier research [46,51]. Regarding the energy use of agricultural machinery, S-III had the highest energy consumption (1397 MJ ha⁻¹), S-I had the lowest energy consumption (1092 MJ ha⁻¹), and S-IV had an energy consumption of 1289 MJ ha⁻¹ (Figure 3b). Irrigation represented the largest share of input energy at 35% in the old lands and 30% in the new lands.

The highest labor energy was in S-I (1051 MJ ha⁻¹) and the lowest was in S-IV (498 MJ ha⁻¹) (Figure 3c). Regarding the organic fertilization process, the S-I and S-II scenarios had the highest labor consumption with a value of 187 MJ ha⁻¹, because organic fertilization of the crop was performed by broadcasting (manual), while the S-III and S-IV scenarios were the lowest in labor consumption, with an average of 5 MJ ha⁻¹, due to the use of mechanization in the organic fertilization. There is no noticeable change in the use of workers in plowing because they use the chisel plow in preparing the soil for cultivation in all scenarios. As for sowing, the scenarios of the old lands consume labor energy with a value of 261 MJ ha⁻¹, because they all depend on sowing by hand. As for the S-IV scenario, sowing consumes the least labor energy, with a value of 2 MJ ha⁻¹, because of using mechanization in the cultivation process. In all scenarios, fertilization consumes a small amount of labor energy, valued at 5 MJ ha⁻¹, because farmers spread fertilizers while carrying out the irrigation process, which does not consume much time. Spraying was carried out by knapsack or spray motor in all scenarios, where the average spraying energy was about 2 MJ ha⁻¹. Harvesting was done by a simple harvester in S-II, S-III, and S-IV, while in S-I, the harvesting was accomplished by a hand sickle, making it labor-intensive with a value of 373 MJ ha⁻¹. In all scenarios, a tractor-mounted thresher is used. The maximum threshing energy was 131 MJ ha⁻¹ in S-I, S-II, and S-III, while the minimum was 126 MJ ha⁻¹ in S-IV. As for irrigation, it needs a little more labor energy in the old lands, because it depends on flood irrigation. As for most of the newly reclaimed lands, farmers depend on modern irrigation, which maximizes the efficiency of using water and fertilizers.

3.3. The Indicators of Energy for the Wheat Production Scenarios

The results of the energy indicators, which are energy productivity (kg MJ⁻¹), specific energy (MJ kg⁻¹), energy use efficiency, and net energy (MJ ha⁻¹), are shown in Figure 4. The energy use efficiency on average was 3.399 for all scenarios in the wheat production system in the Egyptian Delta. In other studies, the energy use efficiency of wheat production on average was 3.00 [16], 6.94 [52], and 1.48 [7]. Low energy use efficiency is evidence of resource mismanagement in wheat cultivation [53]. S-III gave the highest energy use efficiency (3.770) while S-IV had 3.475. The wheat production per hectare in the old lands was more than in the new lands; despite this, the values were close because the energy inputs were low for S-IV. In comparison, the energy use efficiency was low in S-I (3.040) due to lower productivity and higher energy input.

The results also show that an average yield of 1 kg of wheat requires 6.633 MJ with an energy productivity of 0.152 kg MJ⁻¹. The production energy requirement of 1 kg of wheat was the lowest in S-III, with a value of 5.768 MJ ha⁻¹ and an energy productivity of 0.173 kg MJ⁻¹, while the highest energy-consuming scenario for producing 1 kg of wheat was S-I at 7.499 MJ kg⁻¹. This means that S-III is the best-case scenario in old lands farming because every 1 MJ gives 0.173 kg of wheat grain compared with S-I and S-II. This is due to using mechanization in farming operations, which increases productivity and reduces input resources [50]. S-IV outperformed the S-I and S-II scenarios for the old lands, which had 6.431 MJ kg⁻¹ for specific energy, and this result is good for the farmer in the new, unfertilized land. The net energy is a critical indicator of the input-output energy gap. The highest net energy in old lands in the Egyptian Delta was S-III (117,862 MJ ha⁻¹), and the lowest was S-I (93,598 MJ ha⁻¹); the cause of this is that farmers use mechanization in most of their agricultural operations. The net energy of S-IV was less than all scenarios for the old lands, at 93,006 MJ ha⁻¹; the reason for this is that the yield of wheat in the newly reclaimed lands was less than that of the old lands.

3.4. Optimum Energy Savings and Energy Requirements for Wheat Production in the Old Delta Lands

The DEA was used to investigate the energy consumption pattern in the old Delta lands. As shown in

Table 5. The technical efficiency of the scenarios used in the old lands in the Delta of Egypt.

DMU No.	DMU Name	Input-Oriented CRS Efficiency	RTS
1	S-I	0.83	Increasing
2	S-II	0.92	Increasing
3	S-III	1.00	Constant

, S-III was technically effective, while S-I and S-II were ineffective. The S-III case is on the frontier line. It exhibits 100% technical efficiency, making it the best production case for converting input sources (such as human labor, chemical fertilizers, fuel for irrigation, manure, seed, chemical poison, and irrigation water) into output items (yield and straw). On the other hand, it is possible to optimize the technically inefficient input situations (i.e., S-I and S-II) located in the borderline’s envelopment zone to produce outputs at the same level.

Table 6. Energy saving target for the wheat production system.

Inputs	S-I				S-II
	Optimum Energy Requirements (MJ ha−1)	Saving Energy (MJ ha−1)	ESTR (%)	Share (%)	Optimum Energy Requirements (MJ ha−1)	Saving Energy (MJ ha−1)	ESTR (%)	Share (%)
1. Human labor	433	619	59	7	459	228	33	4
2. Chemical fertilizers	12,065	4616	28	49	12,804	2732	18	51
3. Fuel for Irrigation	5434	1602	23	17	5767	1047	15	19
4. Manure	670	201	23	2	711	124	15	2
5. Seed	3162	666	17	7	3355	329	9	6
6. Chemical poison	1335	361	21	4	1417	128	8	2
7. Water	5720	1378	19	15	6071	787	11	15
Total energy	28,819	9441	25	100	30,584	5375	15	100

displays how the input energy sources were optimized in these scenarios. The high energy input case, S-I, has the greatest improvement potential with reductions in all inputs. A moderate energy input case, S-II, has the most moderate improvement potential, with reductions in all inputs. For the farmers in the study area to produce at the same output level, these input improvements can make these scenarios more efficient. The optimal energy savings for wheat production are based on CCR model results. The total energy savings were calculated as 9441 and 5375 MJ ha⁻¹ for S-I and S-II, respectively. As shown in Table 6, chemical fertilizers contributed the most to the overall energy savings in S-I and S-II, with 49% and 51% of the total energy savings, respectively.

Our results concurred with those of Chauhan et al. [54], who reported that the energy inputs of fertilizers and diesel fuel contributed 33% and 24%, respectively, to the total energy savings in paddy production. Also, our results agree with those of Ilahi et al. [7], who described that the energy inputs of fertilizers and diesel fuel contributed 56.5% and 10%, respectively, to the total energy savings in wheat production. In contrast, Nabavi-Pelesaraei et al. [11] reported that the largest input energy item was water (30.99%), followed by fertilizers (24.99%), and fuel (19.52%).

The main cause of water waste and excessive diesel fuel use was improper irrigation techniques. Also, the lack of standard machines for small areas led to increased consumption of chemical fertilizers in the study area. In addition, the areas are very small; they lack appropriate machines to perform the various agricultural operations of such areas. Farmers resort to performing most agricultural operations with manual labor, leading to waste in manure and chemical fertilizers and loss of crops at harvest. As a result, it is advised to choose the proper water distribution in the farms, which results in a decrease in the consumption of fuel and water, the importation of standard and appropriate machinery for small areas, prompt maintenance, and a decrease in the use of chemical fertilizers (mainly nitrogen).

3.5. GHG Emissions Analysis

Table 7. The GHG emissions for the wheat production system.

Input	GHG Emissions (kg CO2-eq ha−1)					GHG Emissions of Target Inputs Units by DEA	GHG Emissions of Target Inputs Units by DEA	GHG Emissions (kg CO2-eq ton−1)					GHG Emissions of Target Inputs Units by DEA	GHG Emissions of Target Inputs Units by DEA
	S-I	S-II	S-III	S-IV	Average	S-I	S-II	S-I	S-II	S-III	S-IV	Average	S-I	S-II
1. Machinery	77.5	95.9	99.2	91.6	91.0	77.5	95.9	12.7	14.7	13.4	15.7	14.1	12.7	14.7
2. Diesel fuel	665.1	695.2	699.2	603.3	665.7	586.6	644.0	108.7	106.2	94.8	103.3	103.2	95.9	98.4
3. Chemical fertilizers
(a) Nitrogen (N)	309.4	287.7	256.8	232.1	271.5	223.3	237.0	50.6	44.0	34.8	39.7	42.3	36.5	36.2
(b) Phosphorus (P)	11.9	11.4	10.5	9.5	10.8	9.1	9.7	1.9	1.7	1.4	1.6	1.7	1.5	1.5
(c) Potassium (K)	2.7	2.5	2.1	1.8	2.3	1.9	2.0	0.4	0.4	0.3	0.3	0.4	0.3	0.3
4. Manure	365.9	350.9	323.9	314.9	338.9	281.6	298.8	59.8	53.6	43.9	53.9	52.8	46.0	45.7
5. Chemical poison
(a) Herbicide	24.0	22.5	21.7	19.5	21.9	18.9	20.1	3.9	3.4	2.9	3.3	3.4	3.1	3.1
(b) Pesticides	18.2	17.0	16.4	14.6	16.5	14.2	15.1	3.0	2.6	2.2	2.5	2.6	2.3	2.3
(c) Fungicide	3.3	3.3	3.2	3.1	3.2	2.7	2.9	0.5	0.5	0.4	0.5	0.5	0.4	0.4
Total	1478.0	1486.5	1432.9	1290.2	1421.9	1215.9	1325.4	241.5	227.1	194.2	220.8	220.9	198.7	202.5

shows GHG emissions for all scenarios before and after applying DEA to the wheat production system, including machinery, diesel fuel, manure, chemical fertilizers, and chemical poison. The total GHG emissions were the highest in S-II at 1486.5 kg CO_2-eq ha⁻¹, followed by S-I and S-III at 1478.0 and 1432.9 kg CO_2-eq ha⁻¹, respectively. After applying DEA, the total GHG emissions of S-I and S-II were 1215.9 and 1325.4 kg CO_2-eq ha⁻¹, respectively. Thus, such improvements in inputs can make these scenarios more efficient and should be adopted by the farmers in the study area. The new lands scenario had the lowest GHG emissions at 1290.2 kg CO_2-eq ha⁻¹ due to the use of agricultural mechanization reducing the amounts of other inputs.

The highest cause of GHG emissions was diesel fuel for all scenarios, and the highest value was 699.2 kg CO_2-eq ha⁻¹ in S-III. In contrast, the lowest GHG emissions of diesel fuel was in S-I (665.1 kg CO_2-eq ha⁻¹) in the old lands, depending on the level of mechanization. Other input sources were minor contributors to GHG emissions, in agreement with the literature [16,46,51,55]. One ton of wheat grain yield production released 241.5, 227.1, 194.2, and 220.8 kg CO_2-eq ton⁻¹ for S-I, S-II, S-III, and S-IV, respectively.

After applying DEA, one ton of wheat grain yield production released 198.7 and 202.5 kg CO_2-eq ton⁻¹ for S-I and S-II, respectively. For all scenarios, the largest share of GHG emissions was diesel fuel for machinery and irrigation, with 45.0, 46.8, 48.8, and 46.8% for S-I, S-II, S-III, and S-IV, respectively (Figure 5). For old lands, diesel fuel used for machinery and irrigation in S-III was the highest due to the use of agricultural mechanization in most agricultural operations. Also, S-II and S-III consumed more fuel energy than S-IV due to the increasing irrigation fuel amount in old lands than in the new lands.

4. Conclusions

Crop production may be sustained by measuring the overall energy of crops, energy productivity, and energy usage efficiency, which improves energy efficiency and reduces GHG emissions. This study estimated GHG emissions and input–output energy analyses in small-farm wheat production scenarios in Egypt. The results showed that S-III in the old lands had the lowest input energy consumption (42,555 MJ ha⁻¹) and the largest output energy (160,418 MJ ha⁻¹), with an energy use efficiency of 3.77. S-III had the largest output energy due to the use of agricultural machinery in most operations. On the other hand, the total input and output energies in newly reclaimed lands were lower than those in the old lands, which were 37,575 and 130,581 MJ ha⁻¹, respectively, because of less fertilization at the newly reclaimed lands. The energy use efficiency of S-IV was 3.475. The results showed that the highest net energy in the old lands of the Egyptian Delta was 117,862 MJ ha⁻¹ for S-III, and the lowest value was 93,598 MJ ha⁻¹ for S-I. S-IV was lower in net energy than all scenarios for old lands (93,006 MJ ha⁻¹). The GHG emissions were highest in S-II with a value of 1486.5 kg CO_2-eq ha⁻¹ and were lowest in S-III with a value of 1432.9 kg CO_2-eq ha⁻¹. Due to the use of mechanization, the new lands scenario (S-IV) showed the lowest GHG emissions (1290.2 kg CO_2-eq ha⁻¹). For the energy balance for wheat and to identify potential energy losses, the data were analyzed by DEA, which provided us with optimal requirements and energy-saving targets. The results showed that the total energy savings were calculated as 9441 and 5375 MJ ha⁻¹ for the S-I and S-II scenarios, respectively; it was noticed that chemical fertilizers contributed the most to the overall energy savings, followed by fuel for irrigation. The small farms in the Egyptian Delta lack appropriate mechanization to perform the various agricultural operations of such areas. Farmers resort to performing most agricultural operations with manual labor, leading to waste in manure and chemical fertilizers and loss of crops at harvesting. Also, the main cause of water waste and excessive diesel fuel use was improper irrigation techniques. As a result, choosing the proper water distribution in the farms is advised, which decreases fuel and water consumption. Using standard and appropriate machinery for small farms increases wheat production and energy resource saving. For example, the use of solar energy in irrigation can be a useful way to reduce fuel consumption in irrigation, and this also has a major role in reducing the percentage of GHG emissions.