Characterization of capacitive sensors for measurements of the moisture in irrigated soils

Characterization of capacitive sensors for measurements of the moisture in irrigated soils

A. Rende^I; M. Biage^II

^IFederal University of Uberlândia Department of Mechanical Engineering UFU – FEMEC. E-mail: [e-mail] ^IIFederal University of Uberlândia Department of Mechanical Engineering UFU – FEMEC. E-mail: [e-mail]

ABSTRACT

The irrigation is a technique developed to supply the hydric needs of the plants. The use of the water should be optimized so that the culture just has enough for its growth, avoiding waste. The objective of this work was to characterize the behavior of capacitive sensors of humidity to monitor the moisture in the soils. In first instance, it was appraised sensors with dielectric built of synthetic pomes stone (Rd = 0,4 and Rd = 0,8) and of soil samples (Rd = 0,8 and Rd = 1,0), being the Rd parameter a geometric factor that relates the distance between the capacitor plates with radius of the plates. For the calibration, the sensors were installed in PVC recipient of cylindrical shape, filled with soil. The set (sensor and soil) was humidified by capillary effect and submitted by a natural drying very slowly. The parameter readings were taken daily, which allowed obtain the curves relating the humidity percentage, expressed in terms of dry weight, with the output voltage fort the sensor. The experiments were performed in sand soil and in dark red latossolo. The obtained results allowed to infer that the behavior of the sensor has a specific feature for each type of soil, being, therefore, necessary to develop a own calibration curve for the sensor, when used in soil with specific characteristic.

Keywords: Automation, agricultural instrumentation, capacitive sensor

Introduction

The constant elevation of the costs of agricultural production, the needs of agricultural products in larger amount in each time and the needs to have compatible prices with the consuming market, come doing with that the man develops means and equipment adapted for an efficient and profitable agriculture. The irrigation is one of those developed techniques. The primary objective of irrigation is to apply water to maintain crop evapotranspiration when the precipitation is insufficient and water store in the soil has been depleted below a level which decreases crop productivity significantly.

The agricultural exploration in irrigated areas permits a larger productivity when compared with to the dry areas (Siqueira et al, 1987). In addition, the irrigation minimize the characteristics problems of the agriculture totally climate dependent (Spinoza and Lins, 1986). The water deficiency is usually the factor more limited to obtain high productivity (Marouelli, 1996). According to Cardoso (1998), the products obtained in these areas are of better quality and they possess a larger commercial value. Nowadays, the area irrigated in the world corresponds to 17% of the total areas cultivated and it is responsible for 40% of the total production of foods. The Brazil demonstrates a great potential in the expansion of its irrigated agriculture. At present just 5% of its cultivated area is irrigated, corresponding to 16% of the total agricultural production and for 35% of the financial value of that production (Santos, 1998).

In spite of the great benefits provided by the irrigation, it exist evidences that large irrigated areas is becoming inappropriate to the agriculture along the time. The irrigation has been causing negative effects to the soil, and to the quality and amount of available water. Others negative aspects of the irrigation inappropriate use, are that it affect the health of the men, cause severe problems to the fauna and the flora, and in some cases, it affects the socioeconomic conditions of the local population. In order, to minimize the effect of these problems, agriculture must make more efficient use of its irrigation water by putting the proper amount of water on crops at the proper time in order to optimize crop productivity and protect the environment (Bernardo, 1998).

How much water to be applied and when to irrigate are the two basic questions of irrigation system management (Camargo and Pereira, 1990). The determination of the amount of necessary water is defined according to each culture type and to appropriate period of the plant development and controlled by the local climatic conditions (Bernardo, 1995). If this is not considered, the plant growth and the production are severely affected.

In general, irrigation systems management fall into three categories: measurements of the plant and soil components and real time measurements of meteorological variables. Combinations of the three categories are sometimes used. Usualy, the data obtained from the soil, are used to indicate the irrigation moment and the data obtained either from the climate or the plant determine the amount of water to be applied (Andrade et al, 1998).

The climatic elements (atmospheric temperature, solar radiation, etc.) can be used in the determination of the hydric needs of the plants; however, usually the evapotranspiration is used to evaluate the hydric needs. The evapotranspiration rate defined by Sediyama et al. (1998), as being the amount of water transpired by the plant plus the water evaporated directly of the soil surface or of the plant surface.

In the evapotranspiration process, meteorological factors control the strength of the "sink", soil factor control the source of water and plant factors control the transmission of water from the source to the sink. A monitoring and control system based on feedback real time measurement of meteorological variables could be used to maximize water use efficiency and productivity (Gomide, 1998).

The direct measure of some plant factors have been also used in the irrigation programming, i. e., in the determination of the water amount to be applied and in determination of the irrigation timing. Among the plant factors used, the ones that possess larger importance are the measure of the stem diameter, the leaf water potential, stomatal resistance and the plant reflectance.

For the purpose of irrigation, soil water content may be determined by direct or indirect methods. Direct methods are those whereby water is extracted from the sample by some external force and the amount of water extracted is measured. Indirect methods are those whereby certain properties of the soil are affected by is water content and the effect of the water on this property is measured; sometimes, the measurement is performed on object (sensor) which is installed in the soil and which is assumed to be at some water content equilibrium with the surrounding soil. However, the water content of the sensor (usually a porous body) at equilibrium will depend on the energy status of the soil rather than its water content.

Indirect measurement of soil water content was proposed nearly 100 years ago using electrical resistance measurement between two parallel electrodes embedded in soil Later, the parallel electrodes were embedded in gypsum (Reichardt, 1985). When permeated with water, the electrical conductivity of gypsum approximates that of an average texture soil at the same water content. At equilibrium with the surrounding fluid, the electrical resistance of the block decreases with increasing water potential. Commercial gypsum blocks are porous enough to permit a small amount of calcium sulfate (a weak electrolyte) to dissolve with the permeating water, thereby creating a relatively buffered, stable conducting medium. The measured voltage output is a function of the electrical resistance and therefore of the matric potential with which it is at equilibrium. However, the relationship between voltage output and humidity present in the soil is strongly influenced by the chemical concentration of the soil, consequently, by water immerse in the soil and the sensor. In the same way, this relationship is also strongly influenced by the soil temperature. Along the various years of studies, several researchers have been tried to minimize these influences on the Electrical Resistance Sensor, however, the success was been limited. Among these researcher, it can be referenced the works of Daker (1983), Reichardt (1985) and Armstrong et al. (1985 and 1986).

Additionally, electrical capacitive sensor can be used to indirect measurement of soil water content. Recently, the performance of this instrument and its usefulness in the laboratory and in the field tested and evaluated (Gomide, 1998, Reichardt, 1985).

This work has the aim to characterize the behavior of capacitive sensors of humidity, built with different dielectric and dimensions, for the monitoring of the moisture in irrigated soils. The choice of a capacitive sensor is due to the fact that the influence of the soil temperature and chemical concentration of the water present in the soil, are very lower and it possesses also a response time relatively fast. Another fact that motivated the performed of this work is the lack of research conducted to improve the technology of the agriculture, which the demand will increase considerably in the future due the augmentation of the need of food production.

Capacitive Sensor of Humidity

Operation Principle

The electrical capacitance has been used largely to characterize the variation of physical parameters in the time. In the case of the use of capacitive sensor, it is desirable to have a two-dimensional electrical field, which simplifies the mathematical model for the sensor. The two-dimensional model is valid only if the distance between the internal surfaces of the capacitor plates, d, is smaller than the capacitor plates areas, S. In this way, the electrical capacitance of the sensor can be defined as being (Pinto, 1997):

where e is the electrical permittivity of the medium.

Through Eq. (1), it is possible to infer that the capacitance depends of the dielectric existing between the capacitor plates and on the capacitor dimensions. In this work, the variation of the medium electrical permittivity, e, is the variable responsible by the capacitance change along the time, once the distance between the capacitor plates, d, and the capacitor plate areas, S, are constants. It can observe in Eq. (1) that the value of the sensor capacitance varies lineally with the electrical permittivity of the medium. Therefore, it was used the variation of the dielectric between the capacitor plates, which is a function of the water amount present in the not sealed porous medium of the sensor was used to monitor the variation of the soil moisture. Thus, the variation of the capacitance depends exclusively on the kind and of the matter amount present between the plates.

In this way, as the soil around the sensor becomes more humid, the porous medium that composes the dielectric absorbs a certain amount of water, proportional to the moisture present in the soil substratum. Similarly, when the soil becomes drier, the dielectric loses water, also as a function of the moisture present around the sensor. In the soil, the hydric motion is characterized by the flow of water from a saturated area (high potential) to a dry area (low potential).

Other point to be considered is the fact that the capacitive sensor is part of an electric circuit, where the capacitance variations due to the change of the water amount present in the soil, should be reflected in values of voltage output. This sensor, which has its impedance varying proportionally to the parameter to be measured, needs a scheme of voltage division, so that its variation can be detected. In all the experiments performed with the sensors, the value of the voltage divisor was fixed as 470 kW. It was a similar electric circuit of one set up by Pinto (1997) that is presented in Fig 1.

The circuit that constitute the capacitive sensor (Fig. 1) was supplied with alternate current (CA) with a sinusoidal wave form, exciting with a frequency of 2 kHz and a amplitude of 9 volts. The choice of this frequency was due it represent the maximum excitation frequency for the sensor, which did not occur amplitude attenuation of the exciting sign, when the sensor is maintained in constant humidity. Generally, in impedance sensors (capacitive or conductive sensors) is necessary to supply the circuit with alternate current with higher frequency to eliminate the load transfer between the electrode solid surface (capacitor plates) and liquid surface in contact.

This phenomenon, denominated of double layer, is characterized by a transfer of electrons between the two ion layers. The electron transfer happens in between the electrode surface and the liquid film in contact with the plate surfaces (Coney, 1973). When voltage is applied, the exciting electric current flowing through the sensor will be a function of the liquid resistance and the capacitive effect produced by the double layer. The liquid resistance is a function of the liquid present inside the medium porous of the sensor. In practice, the equivalent electric circuit can be represented by a serial connection of a capacitor, C_S and resistor R_S (Fig. 1), which products an impedance, I_S, described as

In this case, the excitation frequency influences the humidity measurements present inside the sensor. The excitation frequency has to be low enough, so that the resistance produced by the

Liquid present inside the sensor may be negligible compared to the capacitive impedance, or vice-verse. Moreover, the electrochemical effects of the double layer must be considered. These effects are minimized when the exciting signal is alternated and the signal frequency is boosted by high frequency. This physical aspect may be understood with aid of the Eq. 2.

Fig.2 show the average output voltage, V_S, supplied by the circuit presented in Fig. 1, as a function of the excitement frequency, F _fonte. The circuit was supplied by a sinusoidal source with amplitude of 9 volts. In the results shown in the graph of the Fig. 2 can be observed that the output voltage remains constant for a bandwidth comprise in the range 0 < F _fonte < 2000Hz. For exciting frequencies higher than 2000Hz, a output voltage damping, that is more intense for higher frequencies. However, the sensor capacitance must remain constant, since the humidity presented in the sensor is constant. Thus, the change of the output voltage supplied by circuit for high frequencies represent an increasing in importance of the liquid resistance present in the sensor, since the exciting electric current flowing through the sensor is a function of the liquid resistance and the capacitive effect produced by the double layer. Therefore, due the need to excite circuit with an alternate current with high frequency to minimize the electrochemical effects of the double layer and the need of the sensor working as capacitive (in this case is necessary minimizes the effect of the resistance of the double layer), it was chosen the largest exciting frequency, which did not occur attenuation in the amplitude (increasing the resistive effect of the double layer) and it is high enough to minimize the electrochemical effect ( i. e., 2000 Hz).

However, a better understanding about the behavior of the circuit shown in Fig. 1, it can be obtained through the analytical calculation of the transfer function of the circuit (Eq. 3).

where S = wj and t_c = R_m(C_s + C_pm) is the constant of time that characterizes the answer of the system and w it is the angular speed of excitement of the sensor. C_pm is the capacitance on the porous medium that maintains constant and C_s is the capacitance on the liquid inside porous medium of the sensor. Thus, C_s changes with the moisture inside sensor. Therefore, the sensor is sensitive only a function of C_s.

In the approach refereed in Eq. (3) was considered that R_m >> R_s, what allows to characterize the system as a capacitive sensor.

Therefore, when analyzing the transfer function given by Eq. (3), it is concluded that the system is a pass-low filter of first order.

Construction Scheme of the Sensor

To perform the experiments about the behavior of the sensors, different types of sensors were built. The electrodes of the sensors were built in cylindrical shape of copper foils, with 2 mm of thickness. The plate diameters of the electrodes were of different sizes, with aim to constitute sensors of different ratio of distance between of the electrodes and the electrode radii. It was used as either porous medium synthetic pomes stones or soil samples, constituting the dielectric of the sensor. The synthetic pomes stone used is commonly found in the cosmetic shops. According to the manufacture, it is constituted of water, cement, whitewash sands and powder of aluminum. In the case where the medium porous of the sensor was built with soil sample, it was used compacted soil with different texture.

The copper plates of the electrodes were covered with a protecting varnish. This coating had two purposes. The first, to avoid the oxidation of the copper plates in contact with the soil. The second, and more important, to eliminate effect of electric load conduction through the dielectric. It can occur depending of the water conductivity in the soil. Then, the varnish coating may conduct the sensor to have only a capacitive behavior, since R_sensor® ¥.

The sensor who has the medium porous of the dielectric been of synthetic pomes stone, was built according to the scheme showing in Fig. 3. The sensors whose the medium porous of the dielectric was of sample soil were built in a different form, according to the scheme showed in the Fig. 4. In this case, the electrodes of copper plates were linked by tecnil segments (Fig. 4a). In function of electrodes plate's diameters, a different number of tecnil segments were used to fix the sensor structure (Figs. 4.b and 4.c). Finally, in these sensors, their lateral surfaces were closed by a synthetic mesh, spaced of 2 mm. Table 1 show the geometrical parameters and the dielectric types that constitute the sensors, who is the behaviors, were analyzed in this study. In this table, the R_d parameter is defined as being the ratio of the distance between the plates of the electrodes, d, and the radius of plates of the electrodes, r. It is described by Eq. 4.

The R_d parameter allows analyzing the influence of the geometry on the behavior of the capacitive sensor. Thus, the variations established in the dimensions of the sensor have as aim to obtain results that allow appraising the influences of the geometric aspect on the behavior of the sensors.

It was built sensors with two kinds of materials for the porous medium with aim to analyze the influence of the interaction between the sensor and the soil in study. In reality, there is different porosity for the soil in study and for the sensor porous medium. Nevertheless, there is also different capillary force acting in the sensor and in the soil, thus, it can be expected a non-equilibrium of the humidity between soil and sensor. This physical aspect is very difficult to verify in practice due the difficult to measure the soil porosity and to adapting a sensor with an appropriate porosity for each kind of soil. Therefore, to verify the influence this porous medium characteristics on the sensor behavior, it was study two kind of sensor. The first one has the porous medium with similar material of the soil and another has porous medium with different capillary property of the soil, what allows to quantify the influence the porous medium property.

Characterization of the Humidity Sensor

Evaluation Tests of the Sensors in Laboratory

To characterize the sensor behavior in different operation conditions, it was carried out experiments in three soils with different textures. Table 2 shows the granular composition of the soils used.

In the Table 2, A soil is sorted as sand soil, B soil is sorted as dark red latossolo of clay texture and C is sorted as dark red latossolo of very clay texture.

The step of the soil preparation consisted to dry the soil samples in stove (105ºC < T_s < 110ºC,where T_s is the stove temperature), for a minimum period of 24 hours. This condition is necessary to reach constant weight for the soil mass. After the drying and cooling the soil sample, it was powdered mechanically, sifted in a sieve with 2 mm of mesh. This procedure was adopted in all experiments performed with the capacitive sensor.

For the sensor calibrations, they were installed in PVC recipient of cylindrical shape, filled with soil. The set (sensor and soil) was humidified by capillary effect and submitted by a natural drying very slowly. In the inferior extremity of the PVC recipient were made 5 holes with diameters of 6 mm. These holes allowed the system (soil and sensor) to be humidified by capillarity, for 24 hours. Water accumulated in excess inside the recipient are dropped by the holes in the bottom of the PVC recipient. Additionally, caution was taken to installing the sensors appropriately, completely immerse in the sample soil.

After the saturation condition was attained for the calibration system, a drying process was started in atmospheric conditions in the laboratory, shaded ambient, protected of the wind. The readings of output voltage were daily; being used digital oscilloscopes (Tektronix of 200 MHz sampling). The set (sensor and soil) was also weighed daily, in the same moment of the reading of the output voltage, using an electronic balance, with resolution of 0,01 gram. The determination of the soil moisture was determined by Eq. 5.

where U% is the soil moisture expressed in terms of dry weight, P₁ is the weight of the humid sample soil, P₂ is the weight of the drying sample soil and P₃ is the weight of the recipient. The weight of the drying sample soil was obtained before the beginning of each experiment. In each experiment, the sensor was weighed so that its weight could be subtracted in the results.

With the results of the three soils calibration, were built the curves that represent the sensor behavior, linking the percentage of humidity with the output voltage.

Sensor with Dielectric of Synthetic Pomes Stones

First, it was essayed sensors with dielectric of synthetic pomes stone. To verify the influence of geometrical factors in the behavior of the sensors, experiments were performed with two sensors of different geometrical parameters. The first with 25 mm of diameter and 5 mm of distance between the plates of the electrodes (R_d = 0,4), and another with 50 mm of diameter and 20 mm of distance between the plates of the electrodes (R_d = 0,8). The graphs presented in the Fig. 5 relate the soil moisture with the output voltage of the sensor.

To analyze the results showed in the Fig. 5 it is necessary to introduce some definitions. First, the total potential of water of the soil is defined by the International Soil Science Society as the amount of work that must be done per unit quantity of pure water in order to transport reversibly and isothermally an infinitesimal quantity of water from a pool of pure water at a specified elevation at atmospheric pressure to the soil water. Soil water is subject to a number of force fields such as attraction of the solid matrix for water, the present of solutes, the action of external gas pressure and gravitational which cause its potential to differ from that of pure, free water. The total potential of soil water is defined as the sum of the contributions from these various factors (Eq. 6):

where is the total potential, is the pressure potential, is the gravitational potential, is the matric potential and is the osmotic potential.

Second, the experiments were conducted from the saturation point of the soil, SP, up to values lower than that one corresponding to permanent wilting point, PWP. It will be considered to the analysis of the results three distinct intervals: SP-FC range that comprises the moisture changes in the interval between the SP and the FC; FC-PWP range comprises the moisture changes in the interval between the FC and the PWP. Finally, the measurements, which are the values, are comprised lower than the PWP were negligible, because they remain out of the interesting region.

It is observed in the Fig. 5.a that the curves relating percentage of humidity with output voltage, corresponding to three soils present similar aspects. Therefore, it is verified that the curve obtained in the experiments with sand soil presented smaller changes in the output voltage in the SP-FC range (32,31% < %U < 13,83%) than the FC-PWP range (13,93% < %U < 2,55%). In the PS-FC range has the predominance of the potential pressure regarding to the others components of the total potential of the soil water (Reichardt, 1985). In this analysis is considered that the gravitational potential remains constant, in any situation of sensor operation, because it was always maintained in a same depth level. However, in the FC-PWP range the matric potential acquires larger importance. Considering that the porous material used as dielectric is permeable to the water and salts, thus, the mixture of water and salts contained inside the sensor has the same concentration of the soil. Therefore, it can be affirmed that the variation of the output voltage occurred in the FC-PWP range was not caused by the osmotic potential. Thus, it can be attributed that the variation of the output voltage was exclusively due to the matric potential. This fact agrees with the theory, because the capacitive sensor installed in the soil attains an equilibrium point that is balanced with the matric potential of the soil water and not with the soil moisture.

It was analyzed the sensor behavior working immersed in sand soil with the sensor of R_d = 0,4. However, the others curves presented in the Fig. 5.a, corresponding to the sensor working immersed in dark red latossolo of clay texture and in dark red latossolo of very clay texture, presented qualitatively similar behaviors. In addition, the curves presented in the Fig. 5.a evidenced small variations of the output voltage as a function of the moisture changes in the SP-FC range, in which it was significantly. However, it can be observed also by the curves of this figure large variation on the output voltage as a function of the moisture changes in the FC-PWP range, in which it was moderated. This aspect presented by the sensor behavior became evident that the capacitive sensor is more sensitive to the matric potential than the pressure potential. This is an important conclusion, if it is considered that range of soil moisture change of interesting for irrigation is comprising in the FC-PWP.

Fig. 5.b represents the same calibration results showed in the Fig. 5.a, however, for a capacitive sensor with R_d = 0,8. It can be observed that the two sensors with different geometrical characteristics present qualitatively similar results, when they were immersed in the same kind of soil, but with significant quantitative differences. These aspects were verified for the three soils assayed, essentially, in the SP-FC range. These differences are almost insignificantly in the FC-PWP range for the soils quartozose sand and clay soil, but it was more intense even in FC-PWP range for the very clay soil. Rende (1999) presented the comparison of the curves for the each soil kind, obtained with sensors of different geometrical factors, R_d. The author remarked that smaller geometrical factor, the electrical field of the capacity sensor has more intense a three-dimensional behavior in the sensor edges. In consequence, these edge effects affect the characteristic of the sensor behavior, once this kind of sensor was design to work in a two-dimensional medium. As verified, the three-dimensional effect interfered more significantly when the sensor was immersed in the very clay soil.

The occurrence this fact is attributed to the interaction between the capillary forces that act in soil porous medium and sensor porous medium, since the difference between the capillary forces on the two mediums is larger in very clay soil. This remark is valid for the FC-PWP range, where the matric potential is predominant, which is characterized by the capillary force. Thus, it is comprehensible that when the sensor was immersed in the sand soil, the difference of the curves was minimized, since in this soil kind the capillary force is lower approaching more the capillary force of the sensor.

However, in the SP-FC range the soil water potential is dominated by the pressure potential, where the sensor did not showed sensible enough to measure with accuracy the soil moisture. Thus, the edge three-dimensional effect of the sensor troubled more intense the sensor behavior in this range.

It can be also remarked that the Figs. 5.a and 5.b evidenced changes in the relationship between the output voltage and soil moisture for the different soils. This effect can be attributed to the fact that each soil possesses a characteristic curve of moisture retention that is a physical-hydric property of the soil.

Thus, the moisture retention is a parameter that is a function of the soil texture. In conclusion, in spite of the capacitive sensor suffers little influence either of the soil characteristic or of its chemical composition (Biage, 1998), it is not possible to obtain an universal curve of calibration.

Therefore, the sensor built with dielectric of synthetic pomes stone can be used perfectly to monitor the present humidity in the soil, since it is built the calibration curves that relate the output voltage with the soil moisture. Thus, the indirect measurement of soil water content with the capacitive sensor, allows to estimating irrigation water requirements accurately.

Sensor with Dielectric of Soil Sample

The second group of experiments was leaded with the sensors, of the soil sample dielectric, which were constituted in each experiment by the own soil to be analyzed. For example, in the experiment leads with sand soil, it was used as dielectric of the sensor the same soil. To verify the influence of geometrical factors in the behavior of the sensors, experiments were performed with two sensors of different geometrical parameters: one with 50 mm of diameter and 20 mm of distance between the plates of the electrodes (R_d = 0,8), and another with 60 mm of diameter and 30 mm of distance between the plates of the electrodes (R_d = 1,0). The use of the different geometrical factors, R_d, in this experiments, compared with the sensors of dielectric of synthetic pomes stone experiments, is due to difficult to build smaller sensor with dielectric of soil sample. The graphs presented in the Fig. 6 relating percentage of humidity with sensor output voltage.

In the curves presented in Figs.6.a and 6.b the facts to be remarked are similar of that described when analyzing the Fig.5, relating to the capacitive sensor with dielectric of synthetic pomes stone. Therefore, this kind of capacitive sensor presented similar behaviors regarding to the soil water potential; i. g., being not enough sensible in the SP-FC range, where the pressure potential is predominant and sensible enough in the FC-PWP range, where the matric potential is predominant. In addition, it can be also emphasized that the FC-PWP range is the only interval of soil moisture variation interesting in the irrigation process.

A last fact to be evince related to the sensor behaviors, for both dielectric of synthetic pomes stone and dielectric of soil sample, is the verification of that the sensors with larger dimensions presented larger sensibilities (Table 3). This aspect of the sensor behavior can be verified in the Table 3, where the sensibilities of the sensors embedded in different soils are presented. The evaluation of the sensor carried out, applying the sensibility definition by Eq7:

where S is the sensibility, is the change of the input signal and is the change of the output signal in the electric circuit.

In this work the sensibility of the sensor was evaluated, regarding to input signal as the soil moisture and the output signal as capacitance of sensor, which is a function of the moisture in the soil. The capacitance of the sensor is a physical variable more adequate than the output voltage to estimate the sensor sensibility, because there is a linear relation between moisture and capacitance. This fact occurs, essentially, in the FC-PWP range, where the sensor is sensible with the water matric potential, which characterizes the water retention in the soil, constituting the variable of interest in the irrigation process.

From the output voltage signal, it can obtain the capacitance signal, C, as (where f is the excitation frequency and Z_C is the impedance of the electric circuit). Therefore, the sensor sensibility, S, could be evaluated using the Eq. 8:

It is important to evince that the sensibility estimations were performed for FC-PWP range. The results presented in the Table 3 show clearly that sensor sensibility increases, when the geometrical factor, R_d, as described above.

Finally, it was carried out essays with a tensiometer, classical instrument that have been used extensively for monitoring soil moisture potential directly and for irrigation schedule to measure the soil water potential. Tensiometers measure the soil matric potential (negative pore water pressure) in soils. The experiment was leads in the manner similar of that one explained in the section of the experimental procedure. The only difference in this experiment is that now it was proceeded measurements simultaneously with the capacitive sensor and with the tensiometer. The results are presented in the Fig. 7. It can observed in this figure that the curves obtained with the capacitive sensor and with the tensiometer, immersed in soil sand, agree satisfactory, with a relative error less than 3%, for the two sensors studied. This fact evinces in a convincing way that the capacitive sensor measures the soil matric potential in soil.

Conclusions

In this study, it was analyzed two kinds of sensors: one with dielectric of synthetic pomes stone and of soil samples of soil, with different geometrical dimensions. It was observed that all sensors calibrated enter in balance with the soil matric potential in soil, and not with the presence of the water in the soil. This fact is of large importance, since the theories that establish approaches of irrigation and determine the amount of water consumed by the plants, are based on the quantification of soil matric potential. Inside of this context, it can be said that this verification constituted one of the most important aspects of the performed study. However, this fact has already verified and described in the literature, but with few studies evidencing in a convincing way this aspect, mainly, for capacitive sensor.

Another important aspect that the results allowed to infer is that the sensor capacitive has characteristic behavior for each soil type, being therefore, necessary to develop calibration curves for each sensor in use.

• Andrade, C. L. T. de, "Parâmetros de solo-água para a engenharia de irrigação e ambiental", Anais do XXVII Congresso Brasileiro de Engenharia Agrícola, Poços de Caldas, MG, Brasil, pp. 1-46, agosto 1998.
• Armstrong, C.F., and J. T. Ligon. "Moisture uptake patterns of non-irrigated and trickle-irrigated peach roots." ASAE paper No. 86-2126. St. Joseph, MI: ASAE. 1986.
• Armstrong, C. F., J. T. Ligon and S. J. Thonsom. "Calibration of watermark 200 soil moisture sensor" ASAE paper No. 85-2077. St Joseph, MI: ASAE, 1985.
• Bernardo. S., "Irrigação e produtividade", Anais do XXVII Congresso Brasileiro de Engenharia Agrícola, Poços de Caldas, MG, Brasil, pp. 117-132, agosto 1998.
• Bernardo, S. "Manual de Irrigação e Drenagem" - 6ª ed., Viçosa: UFV, Imp. Univ., 1995. 657 p.
• Biage, M. "Analysis of impedance probes behavior applied to liquid film thickness measurements", RBCM - Journal of the Braz. Soc. Mechanical Sciences. 20 (4): 485-505, 1998.
• Camargo, A. P., Pereira, A. R., "Prescrição de rega por modelo climatológico". Campinas, Fundação Cargill, 1990. 27p.
• Cardoso, H. E. A., et alli, "As águas da agricultura", Agroanalysis, 18(3): 27-28, março 1998.
• Coney, M. W. E., 1973. "The theory and application of conductance probes for the measurement of liquid film thickness in two – phase flow". J. of Scientific Instruments, Vol. 6, 903-910.
• Daker, A., "A água na agricultura" – 6ª ed.., Rio de Janeiro, Freitas Bastos, 1983, 120 p.
• Gomide, R. L., "Monitoramento para o manejo da irrigação: instrumentação, automação e métodos", Anais do XXVII Congresso Brasileiro de Engenharia Agrícola, Poços de Caldas, MG, Brasil, pp. 133-216, agosto 1998.
• Marouelli, W. A., "Manejo da irrigação em hortaliças", 5ª ed. Brasília: EMBRAPA - SPI, 1996. 72 P.
• Pinto, C. .A. R. "Desenvolvimento de um sistema de monitoramento à distância de umidade de solo". Tese de Mestrado - Faculdade de Engenharia Mecânica de Uberlândia :UFU,1997. 102 p.
• Reichardt, K., " Processos de transferência no sistema solo-planta-atmosfera", - 4ª ed., Campinas, Fundação Cargill, 1985. 445 p.
• Rende, A., "Desenvolvimento de um protótipo para monitorar a umidade do solo: Qualificação do comportamento de um sensor capacitivo". Tese de Mestrado – Faculdade de Engenharia Mecânica de Uberlândia: UFU, 1999. 103 p.
• Santos, J. R. M., "Irrigar é preciso", Agroanalysis 18(3): 29-34, março 1998.
• Sediyama, C. G. et al "Relações clima-água-planta", Anais do XXVII Congresso Brasileiro de Engenharia Agrícola, Poços de Caldas, MG, Brasil, pp. 47-116, agosto 1998.
• Siqueira, P. R., "Efeito da irrigação na produtividade do cacaueiro (Theobroma cacao, L.), Anais do XVI Congresso Brasileiro de Engenharia Agrícola, Vol. I, Jundiaí, SP, Brasil, pp. 116-127, 1987.
• Spinoza, W. G., e LINS, J.F., "A importância da água para a citricultura no estado de São Paulo", Anais do VII Congresso Nacional de Irrigação e Drenagem: 493 - 534, Set 1986.

Publication Dates

• Publication in this collection
  27 Aug 2003
• Date of issue
  July 2002