1. Introduction
Agriculture is one of the most critical activities for human survival and also one of the most vulnerable to climate change [1,2,13]. Irrigation systems play a crucial role in mitigating this vulnerability by reducing dependence on rainfall. Researchers worldwide are actively exploring methods to optimize the use of water resources in irrigation. For instance, Su Ki Ooi et al. [3] developed a tool for designing decentralized controls for irrigation channels. Mareels et al. [4] present a framework for automated irrigation systems using control theory principles, enabling on-demand water delivery and improving overall distribution efficiency.
Given the complexity of hydraulic calculations, personal computers and other technologies are essential tools in this field [5], [6]. This study focuses on achieving uniform flowrate distribution in open-network irrigation systems with fixed pipe diameters. While various methods exist for designing pipelines [7], [8], this work addresses the specific challenge of non-adjustable diameters. Here, using orifices (holes) of varying sizes at each system outlet to achieve uniform flow rates is proposed. Intuitively, the size of the orifice should increase with distance from the water source.
This article presents a methodology for calculating orifice sizes to achieve uniform flow across all outlets. A case study demonstrates the application of the proposed methodology.
2. Methodology
The objective of this work is to determine the size of the orifices that will allow the achievement of the desired flow rates at each outlet of an irrigation system, regardless of whether the flow rates need to be uniform or not. With this in mind, the proposed methodology takes the desired flow rates at each outlet of the system as input data. Therefore, the first step is to apply the continuity equation at each node or branch of the system in order to determine the flow through each section of pipeline.
With the flow rates for the pipelines determined, the head loss can then be calculated using Darcy-Weisbach’s Equation (1).
Where:
f= Frictional resistance (dimensionless).
L= Length of the pipe (m).
V= Velocity of incompressible fluid (m/s).
D= Pipeline diameter (m).
g = Acceleration due to gravity (g = 9.8 m/s2).
h LP = Head loss pipeline (m).
Whenever the flow velocity changes direction or magnitude in a conduit (e.g., at fittings, bends, and other appurtenances), additional turbulence is induced. The energy associated with this turbulence is eventually dissipated into heat and produces a minor head loss, also known as local (or form) loss. The local (minor) loss associated with a particular fitting can be evaluated by Equation 2.
Where:
h LS = Head loss pipeline due to turbulence (m).
K= Loss coefficient for the particular fitting involved (dimensionless).
Once the frictional head losses are known, the total piezometric head at each outlet of the irrigation system can be determined by subtracting the frictional head loss from the piezometric head of the preceding node. To illustrate this, the diagram in Figure 1 is provided.
This way, the total head at each node of the irrigation system can be calculated. The total head at a node is determined by subtracting the head loss (hL) between it and the previous node. This process is repeated iteratively until the total head is calculated for all outlets. The flow velocity at the outlet is directly related to the available piezometric head; a higher piezometric head results in a greater fluid velocity. The outlet flow velocity is then determined by the available total head at that point, as described by Torricelli’s Equation (3), which relates the fluid velocity to the square root of the head.
Where:
V s = Speed of outflow (m/s).
h = Available head (m).
Finally, once the flow rates and velocities at the outlet are known, the outlet hole diameter can be calculated using the well-known equation V = Q / A. Alternatively, this can be achieved by applying the continuity equation between the flow through the pipe and the flow at the outlet hole, and solving for the outlet hole diameter. The diameter of each outlet’s orifice is calculated using Equation (4). This calculation relates the orifice diameter to the flow velocity in the pipeline and the pipeline diameter.
Where:
D0 = Diameter of outlet’s orifice (m).
D p = Pipeline’s diameter (m).
V T = Outflow speed through outlet hole (m/s).
V s = Fluid velocity through main pipe.
This process is repeated at each outlet hole, thus completing the algorithm for calculating the diameters. In this work, the equations resulting from the described formulations are implemented using the Engineering Equation Solver (EES) software, which, as its name suggests, is designed to solve engineering equations.
Figure 2 illustrates the algorithm used to calculate the diameters of the irrigation system’s exit orifices.
If the flow rates aren’t predefined, the algorithm can be used by setting a target flow rate based on empirical knowledge or experience. The target flow rate can be gradually increased in small increments, typically by tenths of a liter per minute, and the maximum value will be the highest one that ensures no zero or negative flow values at any outlet.
The dimensionless friction factor (f) can be estimated for a pipe having surface roughness ∈ (m) using the Colebrook-White equation:
Where:
D p = Pipeline diameter.
Re = Reynolds number.
In Equation (5), the friction factor f appears on both sides of the equation, and there is no explicit solution for it. The calculation of this factor requires an iterative procedure. The EES software provides a built-in function that can be easily called within the code to perform this calculation.
The commonly accepted criteria for Newtonian fluids in laminar flow is that the Reynolds number is less than 2,100, while turbulent flow is characterized by Re exceeding 4,000. However, Freeman [9] suggests a transitional region between 1,225 and 3,000, and Merkle [10] indicates that laminar flow can only re-establish if Re falls below 1,150. Due to the absence of an explicit equation for calculating the friction factor in the transition zone, the approach shown in Equations (5) and (6) is adopted. These equations are plotted in Figure 3 for a material surface roughness of 3×10⁻⁷ m, which is typical for plastic pipes. This assumption is conservative, as it may lead to a slight overestimation of the friction factor and, consequently, the head loss. This approach is preferred, as underestimating the friction factor could result in inadequate water delivery to the farthest outlets of the irrigation system.
3. Cases of study
A theoretical and real-world case study will be presented in this section.
Figure 4 illustrates the structure of the theoretical case, featuring a main pipeline with four branches, each branching into three or four outlets. The lengths of the network depicted in Figure 4 are provided in Table 1. The diameter of all the pipes is 0.0254 m.
Table 1 Dimensions for the theoretical irrigation system.
| Section | Lenght (m) | Section | Length (m) |
|---|---|---|---|
| A-B | 30 | 5-6 | 35 |
| B-C | 25 | 6-7 | 11 |
| C-D | 28 | D-8 | 19 |
| D-E | 27 | 8-9 | 21 |
| B-1 | 16 | 9-10 | 27 |
| 1-2 | 14 | E-11 | 17 |
| 2-3 | 17 | 11-12 | 35 |
| C-4 | 23 | 12-13 | 28 |
| 4-5 | 28 | 13-14 | 11 |
The water inlet is located at point A. The methodology is applied considering various pressures at point A. For this simplified example, each outlet is treated as a single perforation with a specified diameter. All pipes in this theoretical case study have a nominal diameter of 1 inch.
On the other hand, Figure 5 presents a sketch of a real case located in Zacatecas, Zacatecas, México for a fruit tree irrigation system. The methodology outlined in the previous section is applied to this open-channel irrigation network. Table 2 summarizes the required dimensions for defining the network geometry. A water reservoir, located five meters above the irrigation system’s elevation, supplies the water. Head loss due to accessories, including one pipe inlet, six standard elbows, and three globe valves, between the reservoir and the first irrigation system node was also considered. The pipes as well as the accessories are made from PVC.
Table 2 Dimensions for the irrigation system.
| Section | Lenght (m) | Section | Length (m) |
|---|---|---|---|
| A-14 | 0.15 | H-7 | 0.3 |
| A-B | 0.5 | H-8 | 0.3 |
| B-13 | 0.26 | G-6 | 0.3 |
| B-C | 0.87 | G-I | 2.01 |
| C-D | 1.32 | I-5 | 0.3 |
| D-11 | 0.3 | I-J | 2.21 |
| D-12 | 1.45 | J-4 | 0.3 |
| C-E | 1.1 | J-K | 3.08 |
| E-10 | 0.8 | K-3 | 0.21 |
| E-F | 2.62 | K-L | 1.21 |
| F-9 | 0.81 | L-2 | 0.29 |
| F-G | 1.06 | L-1 | 0.71 |
| G-H | 2.17 |
Each outlet was equipped with a set of pipes and fittings, as shown in Figure 6. This set includes a standard elbow, a 30 cm vertical pipe section, a 10 cm horizontal pipe section and a cap with an orifice (diameter to be calculated). All pipes in this real case study have a nominal diameter of ½ inch.
4. Results
The algorithm was tested using the system shown in Figure 4 and the flow rates specified at Table 3. At point A, the system's inlet, a piezometric head of 4 meters is available.
Table 3 Requested caudal distribution.
| Outlet number | Caudal Liters/minute | Outlet number | Caudal Liters/minute |
|---|---|---|---|
| 1 | 0.5 | 8 | 1 |
| 2 | 1 | 9 | 0.5 |
| 3 | 1 | 10 | 1 |
| 4 | 1.5 | 11 | 1.5 |
| 5 | 0.5 | 12 | 1.5 |
| 6 | 1 | 13 | 1.5 |
| 7 | 2 | 14 | 1.5 |
Following the steps of the algorithm, the volumetric flow rates are calculated for each section of the irrigation system using the desired outlet flow rates. This volumetric flow rates are shown in Table 4.
Table 4 Volumetric flow rates.
| Section | Flow rate | Section | Flow rate |
|---|---|---|---|
| A-B | 16 | 5-6 | 3 |
| B-C | 13.5 | 6-7 | 2 |
| C-D | 8.5 | D-8 | 2.5 |
| D-E | 6 | 8-9 | 1.5 |
| B-1 | 2.5 | 9-10 | 1 |
| 1-2 | 2 | E-11 | 6 |
| 2-3 | 1 | 11-12 | 4.5 |
| C-4 | 5 | 12-13 | 3 |
| 4-5 | 3.5 | 13-14 | 1.5 |
Then, the energy losses due to friction are calculated, which are shown in Table 5.
Table 5 Energy losses due to friction.
| Section | Head loss(m) | Section | Head loss (m) |
|---|---|---|---|
| A-B | 1.317 | 5-6 | 0.05831 |
| B-C | 0.7841 | 6-7 | 0.003667 |
| C-D | 0.3528 | D-8 | 0.007916 |
| D-E | 0.1719 | 8-9 | 0.00525 |
| B-1 | 0.00667 | 9-10 | 0.0045 |
| 1-2 | 0.00467 | E-11 | 0.1083 |
| 2-3 | 0.002833 | 11-12 | 0.1274 |
| C-4 | 0.1027 | 12-13 | 0.04665 |
| 4-5 | 0.06271 | 13-14 | 0.00275 |
Using these data, the total piezometric head at each point of interest in the system can be calculated. These results are shown in Table 6.
Table 6 Available head at each outlet.
| Node | Available head | Node | Available head |
|---|---|---|---|
| 1 | 2.676 | 11 | 1.266 |
| 2 | 2.672 | 12 | 1.139 |
| 3 | 2.669 | 13 | 1.136 |
| 4 | 1.796 | 14 | 1.133 |
| 5 | 1.734 | A | 4 |
| 6 | 1.675 | B | 2.683 |
| 7 | 1.672 | C | 1.899 |
| 8 | 1.538 | D | 1.546 |
| 9 | 1.533 | E | 1.374 |
| 10 | 1.529 |
Finally, the diameters of the outlet orifices are calculated based on the desired flow rates and the piezometric heads. These calculations are repeated for piezometric heads at the inlet of 3 m and 2.6 m. The required outlet diameters are presented in Figure 7, where the horizontal axis represents the outlet number and the vertical axis represents the outlet diameter. By applying the proposed methodology, the outlet diameters were calculated. As expected, a larger flow rate requires larger outlet diameters. Additionally, it is observed that the required outlet diameters increase as the available head decreases. This explains why the blue line in Figure 7 diverges from the orange line. In this specific case, a minimum head of 2.6 meters is required to achieve the desired flow rates.
Additionally, the algorithm can be used to determine the maximum uniform flow rate for a given available head at the inlet. This can be achieved by iteratively increasing the desired flow rate at each outlet until a zero or negative value is obtained for the piezometric head at one of the outlets. Figure 8 illustrates the maximum flow rate as a function of the available hydraulic head at point A. This can also be used to calculate the necessary head for a specific uniform flow rate at each outlet. It is important to note that the flow rate does not increase linearly with the head at point A due to the nonlinear increase in pipeline head losses. However, Hower-Cuevas [11] and Echegorri [12] also found a nonlinear relation between flow rate and head loss.
Based on the preceding results, it can be concluded that the model performs accurately and will be valuable for comparing theoretical and experimental results in the subsequent case study. This second case study aimed to achieve a uniform flow rate distribution across all 14 irrigation system outlets (shown in Figure 4), while maximizing the flow rate for each outlet for a 4-meter available head at the inlet. The proposed methodology was applied iteratively, incrementing the flow rate by 0.05 liters per minute, until a negative total head was encountered at an outlet. The maximum flow rate was defined as the highest flow rate with a non-negative total head at all outlets. The calculated outlet diameters for various flow rates are presented in Figure 9. These results reveal the following:
Uniform Outlet Diameters at Low Flow Rates: At low flow rates, the calculated outlet diameters are consistent across all outlets. In this specific case, the maximum achievable uniform diameter is 1 mm, corresponding to a flow rate of 0.35 L/min.
Maximum Achievable Flow Rate: The proposed methodology enables the calculation of the maximum flow rate that can be delivered to each irrigation system outlet. For this specific case, the maximum achievable flow rate is 0.51 L/min.
Diverging Diameters at High Flow Rates: As the targeted flow rate approaches the maximum achievable value, the calculated outlet diameters begin to diverge significantly.
As previously noted, the maximum achievable flow rate is 0.51 L/min. As a result, the flow rate within the pipeline network ranges from 7 L/min in the main feed pipe to 0.51 L/min at each outlet. The calculations also reveal that the main feed pipe accounts for 76% of the total head loss, while each outlet of the system contributes only 0.1% of the pressure drop. Table 7 summarizes the calculated and implemented orifice diameters, along with the achieved flow rates.
Table 7 Obtained flow rate.
| Outlet number | Calculated diameter (mm) | Actual diameter (mm) | Flow rate (L/min) |
|---|---|---|---|
| 1 | 2.68 | 2.5 | 0.51 |
| 2 | 2.67 | 2.5 | 0.53 |
| 3 | 2.64 | 2.5 | 0.58 |
| 4 | 2.44 | 2.5 | 0.52 |
| 5 | 2.31 | 2.5 | 0.55 |
| 6 | 2.15 | 2 | 0.52 |
| 7 | 2.16 | 2 | 0.49 |
| 8 | 2.16 | 2 | 0.48 |
| 9 | 2.03 | 2 | 0.52 |
| 10 | 1.81 | 2 | 0.51 |
| 11 | 1.71 | 2 | 0.44 |
| 12 | 1.71 | 2 | 0.45 |
| 13 | 1.64 | 1.5 | 0.43 |
| 14 | 1.56 | 1.5 | 0.44 |
The targeted flow rate in this case study is 0.51 L/min. However, Table 7 shows that only exits 1 and 10 deliver the exact targeted flow rate. The remaining exits deliver flow rates ranging from 0.43 to 0.58 L/min. This discrepancy between the calculated and the achieved flow rates can be attributed to several factors:
Discrepancy between calculated and implemented diameters: Minor differences may exist between the calculated orifice diameters and the actual implemented diameters due to manufacturing tolerances or limitations.
Transitional flow regime: From node 5 onwards, the flow likely enters the transitional regime (2000 < Re < 4000). In this regime, calculating the friction factor and consequently, the head loss becomes more complex due to the unpredictable flow behavior.
Potential impurities: The presence of impurities within the pipeline, such as glue residue, could introduce additional friction and potentially affect the flow rate.
5. Conclusions
The proposed methodology demonstrates its effectiveness in redistributing flow rates in situations where pipe diameters cannot be modified, such as in existing or pre-purchased systems. For low flow rates, uniform orifice diameters can be achieved. However, to maximize flow rates, orifices with varying diameters are required.
Predicting energy losses within the transitional flow regime (2000 < Re < 4000) and the limitations in manufacturing precise orifice diameters contribute to a ± 15% discrepancy between calculated and actual flow rates.
Using Equations (5) and (6) to calculate the friction factor remains appropriate for this application, as the majority of head loss occurs in the initial sections of the network, where turbulent flow dominates.









nueva página del texto (beta)











