INTRODUCTION:

The particular characteristics of pumping stations (PSs) in rice fields associated with the expanse of irrigated areas in Rio Grande do Sul (RS) when compared to other parts of Brazil, and the quantity of water and energy used in this process justify studies to optimize the use of these resources. Performance indices in irrigation areas are used to identify optimal sizing guidelines. ^{ HYPERLINK "#mkp_ref_003" }) conceptualized a performance index as a quantitative measure of a particular aspect of performance of the managing entity or its service level. Thus, each index is a tool for monitoring efficiency (it measures the extent to which resources are optimally used for production) and effectiveness (it measures the extent to which management goals are met). ^{ HYPERLINK "#mkp_ref_005" }) ranked performance indices by their relevance, selectivity, simplicity, low cost, comprehensiveness, stability, traceability, and reasonability. ^{ HYPERLINK "#mkp_ref_012" }) identified performance indices as tools for monitoring the quality of PSs.

^{ HYPERLINK "#mkp_ref_011" }) studied PSs at rice fields on the western border of RS and established an index that relates the potential of a crop area and pumping height, and concluded that the most efficient PS should have a power consumption value of less than 30W ha^{-1} m^{-1}. However, values observed in the aforementioned study varied from 25Wha^{-1} m^{-1} to more than 100W ha^{-1} m^{-1}. ^{ HYPERLINK "#mkp_ref_004" }) proposed the power used by a pumping volume unit (kW m^{-3)} and active energy consumed (kWh) as indices. This method was also used by ^{ HYPERLINK "#mkp_ref_005" }), who suggested the use of indices that associate energy and power, as measured in the area covered by irrigation units and by the volume of pumped water (kWh ha^{-1}, kWh m^{-3}, kW ha^{-1}, kW m^{-3)}. The overall PS efficiency can serve as a performance index because it is determined by the parameters involved in the process (flow rate, geometric height, energy loss, and power demand). ^{ HYPERLINK "#mkp_ref_001" }) proposed a coefficient of global energy efficiency (GEE) that has two components: the efficiency of the equipment of a PS, and its spatial distribution, considering the water supplying distance, geometric height difference, and irrigation area. The current proposal aims to identify whether the cause of low efficiency is related to the equipment or its spatial distribution. The developed coefficient classifies PSs into categories (excellent, good, normal, acceptable, or unacceptable). The International Water Association (IWA) proposed a performance index associating the consumption of energy per unit volume of water pumped with a gauge height of 100m. This index is known as normalized energy consumption (kWh m^{-3}(100m)^{-1)}; allowing the comparison of PS performance at different heights (^{ HYPERLINK "#mkp_ref_003" }). The same authors proposed the use of normalized energy consumption to evaluate public supply systems, and recommended an approximate value of 0.5kWh m^{-3}for every 100m of gauge height.

The lack of reference values that could be used as decision-making criteria, considering the case of energy efficiency of one PS in a rice fields creates insecurity regarding the exchange, replacement or modification of items that make up the electromechanical system of water elevation in operation. The high investment values associated with special pricing tariff structures for irrigation (with reduced costs) do not encourage alterations necessary for optimizing a PS process. It is estimated that the installed power in rice fields in RS is approximately 500MW. Therefore, a lack of reference performance indices does not allow identification of the conditions of PSs in terms of efficient energy use. Considering this drawback, the goal of this research is to propose and classify performance indices (Pis) for water PSs used in the irrigation of rice fields; the developed method will allow the identification of sizing patterns to increase energy efficiency.

MATERIALS AND METHODS:

The area of study included four municipalities on the western border of the state of RS. Total sample set included 45 properties with 160 PSs, of which 48 were located in Uruguaiana, 60 in Itaqui, 47 in Alegrete, and 5 in São Borja. Names of farmers were provided by the region's electric distribution company (AES SUL), which used as inclusion criteria the farmers' interest in participating in the Energy Efficiency Program (Programa de Eficientização Energética), developed by a company under the supervision of the National Electrical Energy Agency (Agencia Nacional de Energia Elétrica, ANEL).

Overall, the selected fields used flood irrigation with a water layer varying between 3 and 7cm, low wall systems, and an irrigation cycle between 80 and 100 days. PSs were triggered by electrical energy by using a tariff system, Horo-Sazonal green, and by adopting a restricted ordinance schedule.

Collection of field data included: a) Measurement of the geometric pumping height, suction of PS, and the respective piping lengths for total topographic station. b) Measurement of the flow rate by using a portable ultrasonic transit time flow meter. c) Measurement of electrical quantities: active power, voltage, and current, measured using a multimeter installed next to the electrical control board, with reading periods between 15 and 30min. d) Verification of the model and brand of the pump and motor, positioning of the pump and motor (horizontal or vertical axis), use of a floating structure, type of electric starter key, and piping material and its diameter.

Total gauge height of the PS analyzed was estimated from equation 1, which was adapted from the Hazen-William equation, and the hydraulic power derived from equation 2.

where TDH is the total dynamic head (m), h_{g} is the geometric height (m); Q is the measured flow rate (m³ s^{-1)}, C is the Hazen-Williams roughness coefficient (here we adopt a value of 90), D is the pipe diameter (m), L is the length of the pipe (m), L_{virtual} is the equivalent virtual length of the parts of the system (m), P_{hyd} is the hydraulic power (kW), andγ is the specific weight of the water (N m^{-3)}.

The overall efficiency of the system was estimated using the relationship between the hydraulic electrical power and the active power by using equation 3.

where P_{active} is the active power (kW), andη_{global} is the overall efficiency of the combination of all the pumping system equipment.

New projects were then developed based on the real data obtained in the field, aiming the reduction of energy use in the pumping water system. Physical conditions of the PS were not altered. Irrigation area, flow rate, geometric heights, and distances (piping length) were maintained. Piping diameter was adjusted to obtain the best option among standard options made of welded steel sheet. Criterion for selecting a suitable diameter was the least total cost over a period of 10 years of PS operation, based on the sum of fixed costs (piping) and operating costs (energy). After determining the optimal diameter, the TDH was re-estimated using equation 1. The average PS operating time, obtained from energy bills made available by the farmers, was standardized at 1,800h to allow comparisons of consumption rates.

Using the calculated TDH and the established flow rate of the project, the pump with the highest efficiency was selected for each of the 160 cases analyzed, among 68 preselected models (which are available for purchase). The motor was then selected using the required power as a criterion; high efficiency four-pole motors were selected. Thus, an optimized PS was obtained for the required pumping conditions, resulting in minimized energy consumption.

Nine performance indices were applied to the 160 PSs studied for the initial conditions (before) and for conditions in which the sets had been optimized through the substitution of piping, pumps, and motors adjusted for the recommended technical situation (after). The performance indices used are described in equations 4 to 11.

where Pi_{O} is a performance index (W ha^{-1}m^{-1)}, P_{active} is the active power (kW), A is the planted area supplied by the PS (ha), h_{g} is the geometric pumping height (m), Pi_{1} is a performance index (kW ha^{-1)}, Pi_{2} is a performance index (kWh ha^{-1)}, ∆t is the duration of pumping (h), Pi_{3} is a performance index (kWh m³), Vol is the volume of water pumped during the entire irrigation cycle, Pi_{4} is a performance index (kW (m³ s^{-1)}
^{-1}m^{-1)}, Q is the flow rate (m³ s^{-1)}, TDH is the total dynamic head (m), Pi_{5} is a performance index (kW (m³ s^{-1)}
^{-1} m^{-1)}, P_{nominal} is the nominal power of the motor (kW), Pi_{6} is a performance index (W ha^{-1}m^{-1)}, and Pi_{7} is a performance index (kWh m^{-3}(100m)^{-1)}.

The indices defined above relate hydraulic, physical, and electrical parameters, thus allowing to identify the operational performance patterns of PSs. Categories of performance were proposed for each index studied, based on the results obtained before and after optimization of PSs (^{ HYPERLINK "#mkp_ref_006" }). Categories were defined based on the amplitude of the values obtained, creating five equidistant classes, which are called "very poor", "poor", "good", "very good", and "excellent".

The relationship between the indices and the measures used for their determination (geometric height, irrigation area, unit flow rate, gauge height, overall production, and load loss) were tested using linear regression analysis.

RESULTS AND DISCUSSION:

Results showed that optimization reduced the standard deviation and amplitude for every index, indicating that adjustments made in relation to the choice of piping, pumps, and motors led to a decrease in their variability. The variability that remained after the process is due to factors affecting the indices and those that were not altered, such as the unit flow rate (L s^{-1} ha^{-1)}, geometric pumping height (m), piping length (m), and efficiency of the selected pump (%),which was limited by options available in the market.

Comparing the Pi_{0} index "before" and "after" (see HYPERLINK "#t1" ), it was observed that its value was adjusted by the process to less than 30W ha^{-1} m^{-1}, as indicated by ^{ HYPERLINK "#mkp_ref_011" }) as the threshold for obtaining good performance. Regarding the Pi_{1} index, even the mean corresponding to the "before" condition (see HYPERLINK "#t1" ) was lower than the value found by ^{ HYPERLINK "#mkp_ref_007" }) of 0.684kW ha^{-1} in 31 PSs in the central region of RS.

Pi_{2} and Pi_{3} indices, representing the consumption of PSs, with means of 798.83kWh ha^{-1} and 0.064kWh m^{-3} respectively, differed from the values obtained in other regions for different irrigation systems, such as those reported in ^{ HYPERLINK "#mkp_ref_010" }) (2,792kWh ha^{-1} and 0.872kWh m^{-3)}, ^{ HYPERLINK "#mkp_ref_014" }) (0.16kWh m^{-3)}, ^{ HYPERLINK "#mkp_ref_013" }) (0.3kWh m^{-3} and 378.33kWh ha^{-1)}, and ^{ HYPERLINK "#mkp_ref_008" }) (0.47kWh m^{-3} and 4,401kWh ha^{-1)}. This result showed the great variability of these indices, which makes difficult to compare different conditions. This variability is associated with geometric heights typical of the area, the irrigation system used, and the type and size of pumps.

Although Pi_{4} and Pi_{5} indices are expressed in the same units of measurement, they represent different measures. Pi_{4} refers to the active power, whereas Pi_{5} indicates the nominal power of the motor. It follows that service factors smaller than one unit could lead to a higher Pi_{5}value than that obtained for active power. However, the determination of Pi_{5} is facilitated by the use of a power value indicated on the motor plate (motor's nominal power). Similarly, the height reference, used for deriving both indices, is different because Pi_{4} refers to the gauge height and Id_{5} refers to the geometric height, what is easy to obtain.

The Pi_{6} index was proposed because of the ease with which data requiring the nominal power of the motor, irrigated area, and geometric pumping height were obtained. This index and Pi_{5} were the indices with the greatest percentage variation between "before" and "after" ( HYPERLINK "#t1" ), according to the adjustment of nominal power that was carried out in relation to the active power.

The mean value for index Pi_{7} obtained after optimization was 0.363kWh m^{-3}(100 m)^{-1}, which differed from that proposed by ^{ HYPERLINK "#mkp_ref_003" }), who recommended a value of approximately 0.5kWh m^{-3} (100 m)^{-1}. The value obtained is due to the shorter piping lengths, when compared to those used in public supply services, on which the author's recommendations are based. This index shows a strong correlation with Pi_{4}, although Pi_{4}t differs in the measures used for its determination. Hence, both indices were retained, in spite of their similarities (Pi_{4}=Pi_{7}× 36).

The initial mean overall efficiency (before) was 57%, a value that was slightly higher than that observed by ^{ HYPERLINK "#mkp_ref_002" }) (45%), ^{ HYPERLINK "#mkp_ref_011" }) (47%), and ^{ HYPERLINK "#mkp_ref_009" }) (50%), caused by the improvement of installations over time. Process of optimization resulted in an increase of 31.2%, thus reaching an average of 75%, a value obtained based on the improved fit of the selected piping, choice of t proper pump, and motor adjusted to the application. This value differs from that obtained by ^{ HYPERLINK "#mkp_ref_001" }); their study classifies levels of global production higher than 50% as "excellent". The variation reported between "before" and "after" shows the economic potential that can be obtained by the adaptation of PSs.

The partial result obtained when the effect of piping substitution alone was studied in relation to active power (kW) showed a reduction of 10.19%, which led to the same reduction in energy consumption and costs. However, piping, pump, and motor replacement allowed an estimated reduction in active power consumption of 31.20%.

The indices analyzed are related to each other and the measurements, as shown in table 2. The Pi_{0} index is related to the unit flow rate (q (L s^{-1} ha^{-1)}), where as Pi_{1}, Pi_{2}, and Pi_{3} indices are related to the geometric height. Pi_{4} and Pi_{7} indices, in contrast, are shown to be strictly dependent on the efficiency of the equipment. The relationships that were considered significant are presented in table 2.

HYPERLINK "#t3" shows the classification proposed for the nine performance indices presented in this study, indicating the range of values corresponding to"excellent", "very good", "good", "poor", and "very poor". It was suggested the use of the following indices: overall efficiency (η_{g}), Pi_{4}, and Pi_{7}, in this order, to verify the performance of projects; Pi_{1}, Pi_{2}, and Pi_{3}, in this order, to verify whether the location's topography is favorable for rice cultivation; Pi_{0}, when comparing stations and identifying the energy consumption due to excessive water use in the irrigation process; and Pi_{5} and Pi_{6} indices, in case of difficulties in measuring active power or performing a quick estimate.

CONCLUSION:

The combination of physical, hydraulic, and electrical parameters of PSs studied allowed the creation of indices that characterize performance in relation to energy consumption for water pumping in irrigated rice fields in RS. Using the overall cost minimization criterion, the choice of piping size, and more efficient pumps and motors resulted in a reduction of 31.20% in the energy consumption associated with the pumping process, relative to the configuration observed in the field for the 160 PSs studied.

For reference values of size or PS operation estimates, it was suggested the use of the following index values to achieve an acceptable index corresponding to a "good" rating: Pi_{0}<40W ha^{-1} m^{-1}, Pi_{1}<0.50kW ha^{-1}, Pi_{2}<750kWh ha^{-1}, Pi_{3}<0.060kWh m^{-3}, Pi_{4}<19.3kW (m^{3} s^{-1)}
^{-1} m^{-1}, Pi_{5}<24.5kW (m³ s^{-1)}
^{-1} m^{-1} (based on the nominal power of the motor), Pi_{6}<44.2W ha^{-1} m^{-1} (based on the nominal power of the motor), Pi_{7}<0.536kWh m^{-3} (100m)^{-1}, and η_{g} (overall efficiency) >50%.