Copy (2) of Elec140704.qxd Vol. 3, No. 1 (2006) 63-68ring Researchrnal of EngineeThe Jou 1. Introduction Oil wells, in general, produce oil, water, and gas. Knowing the percentage of water in oil is very important for oil production. Oil from nearby wells is carried via pipelines into a collection station. Basic Sediment and Water (BS & W) devices are usually used to determine the percentage of the water in the produced oil. Conventional BS & W devices are based on measuring the capacitance or impedance characteristics of the oil (Smit, et al. 1998; Lucas, 1994). However, such devices cannot be used in a full range of water percentage (0-100%). Other devices such as radioactive and microwave (Shaofan, 1994; Beckwith, et al. 1981) are harmful and require a special operation procedure. Fiber optic sensors (Betta, et al, 1993; Giallorenzi, et al. 1986; Grattan, 1987) are also available. However, these devices are expensive. In this paper we are reporting experimental results for a new, user friendly, and accurate off-line device that can determine the water constitute in oil. The experimental results show that the device can attain a very high resolution up to +/- 0.4% and it can be used to measure a full range of water percentage levels (0-100%). In addition, the experimental results were compared with the theoretical ones and showed a good agreement. _______________________________________ *Corresponding author’s e-mail: mhabli@squ.edu.om 2. Proposed Water-Cut Measurement Method The proposed measurement method of the water-cut in a mixture of oil and water is based on the pressure of a sample from the mixture. Consider the system in Fig. 1. The system consists of an apparatus where the pressure of a liquid sample can be measured. The needed sample from a liquid is of a fixed quantity of height Hs. In principle, for a water and oil mixture, the mixture sample will con- sist of heights Ho and Hw, respectively: (1) where, Ho = height/level of water Hw = height/level of oil The pressures of the oil and water are: (2) (3) where, P = pressure, kPa g = gravitational acceleration, m/s2 Inexpensive and Accurate Measuring Device for Water Constitute in Oil M. Habli*1, M. Meribout1, A. Al-Naamany1, K. Al Busaidi2 1 Electrical & Computer Engineering Department, Sultan Qaboos University, P.O. Box 33, Muscat 123, Oman 2 Petroleum Development Oman, P.O. Box 81, Muscat 113, Oman Received 14 July 2004; accepted 24 October 2005 Abstract:This paper presents an inexpensive and accurate measuring device for water constitute in oil. The new device is based on the relationship between the water constitute in oil and the pressure of a sample from the oil. Experimental results show that the device can attain a very high resolution that can reach up +/- 0.4% and it can be used to measure a full range of water percentage levels (0-100%). Experimental results showed good agreement with theory. Keywords: Water percentage, Oil, Pressure, Inexpensive measuring device swo HHH =+ gHP ooo ρ= gHP www ρ= âjõdG »a AɪdG ᫪c ¢SÉ«≤d ≥«bO h …OÉ°üàbG RÉ¡L »∏ÑM .Ω1@•ƒHôe .Ω ^1»fɪ©ædG .CG ,1…~«©°SƒÑdG .∑ ,2 áá°°UUÓÓîîddGGøe ¬æ«Y øe §¨° dG h âjõdG »a AɪdG ᫪c ø«H ábÓ©dG ≈∏Y ~ªà©j ~j~édG RÉ¡édG Gòg .âjõdG »a AɪdG ᫪c áaô©ªd ≥«bOh …OÉ°üàbG ¢SÉ«b RÉ¡L Ω~≤J ábQƒdG √òg : ≈dG áaÉ°V’ÉH .AÉe (%100-0) øe ä’ÉëdG ™«ªéd πª©à°ùj ¿’ É° jG ¬æµªj h +/-%0^4 ≈dG õ««ªàdG ≈∏Y áJQ~b π°üJ ¿G øµªj RÉ¡édG ¿G äô¡XG á«∏ª©ªdG èFÉàædG .âjõdG .√~«L ¬≤aGƒe äô¡XG h ájô¶ædG èFÉàædÉH âfQƒb á«∏ª©ªdG èFÉàædG ¿Éa Gòg áá««MMÉÉààØتªddGG ääGGOOôôØتªddGG.…OÉ°üàbG ,§¨°V ,âjR ,AɪdG áÑ°ùf : 64 Vol. 3, No. 1 (2006) 68-68ring Researchrnal of EngineeThe Jou H = height/level, m (Pressure Head); ρ = density, kg/m3 The subscripts (o, w and s) stand for oil, water and sam- ple, respectively. The total pressure Pt at the bottom of the container is equal to: (4) where Pa is the atmospheric pressure. Subtracting Pa from (4) yields: (5) Ps is the pressure of the mixture sample which reflects the heights or the fractions of both oil and water in the sam- ple. This system can be used to indicate the fractions of water and oil for different mixtures. Substituting Eqs. (2) and (3) in Eq. (5) yields: (6) Equations (1) and (6) are two independent equations with two unknowns, Ho and Hw. The sample quantity Hs in Eq. (1) is known and fixed. On the other hand, the sample pressure Ps in Eq. (6) is the measured pressure. For known ρo and ρw , it is possible to solve Eqs. (1) and (6) for Ho and Hw as: (7) and (8) This device can also be used to find the densities of the oil and water samples ρo and ρw , respectively, from the pressures of two samples of known heights. For example, Eq. (6) can be written for the first sample as: (9) Similarly, for the second sample, Eq. (6) can be writ- ten as: (10) Equations (9) and (10) can be solved for: (11) and (12) It should be noted that this device is an off-line meas- uring instrument. In order to determine the water-cut in a flow, a sample from the flow is needed. This can be taken automatically from a side pipe through a valve that can close and open automatically by a switch or from a base station computer command. From the pressure of a known amount of the flow, we can determine the percent- age of water in oil. 3. Experimental Setup The experimental setup consisted of a piezo-resistive pressure sensor. The sensor was attached to the bottom of an apparatus to measure the pressure of a sample. The out- put of the sensor was connected to an amplification circuit as shown in Fig. 2. The sample height, Hs , for all experi- ments was set to 25 cm. The pressure sensor that was used in the experiments was a differential type, i.e. the output of the sensor indicates Ps as in Eq. (5). Figure 3 shows a photo for the experimental setup. The experimental samples were prepared using tap water and motor oil. 4. Experiments A number of experiments were conducted to test the accuracy and the capability of the device. In the first experiment, a full range test was conducted starting from a 25 cm water level and no oil and ending with no water Glass Tube Hs = Ho + Hw Ho Hw Piezo-resistive Pressure sensor Ps=Po+Pw = ρoHog + ρwHwg Figure 1. A pressure measuring apparatus awot PPPP ++= wos PPP += gHgHP wwoos ρ+ρ= o ww s o H g P H ρ ρ− = osw HHH −= gHgHP 1ww1oo1s ρ+ρ= gHgHP 2ww2oo2s ρ+ρ= 2w 2oo 2s w H H g P ρ− =ρ 2w 2o1w 1o 2s 2w 1w1s o H HH H g P H H g P − − =ρ 65 8 63-6Vol. 3, No. 1 (2006)ring Researchrnal of EngineeThe Jou and a 25 cm oil level with 1 cm increments. Theoretical results were deduced from the equations presented in Section 2.0 and compared with the experimental results. For the theoretical analysis, an oil density of ρo=0.87 gm/cm3 and a water density of ρw=1.104 gm/cm3 were used. Figure 4 shows both the theoretical and experimen- tal results. The data indicate that since oil has a less den- sity, the pressure dropped as the oil level increased. The results show almost a linear relation between the oil level and the voltage. The second experiment was designed to test the accu- racy of the device and if it can detect small changes in the mixture. This experiment began with a 25 cm water level and no oil. The oil in the mixture was then increased from zero to 0.8 cm using 0.1 cm increments. The results of measurements are shown Fig. 5. From these results, it is clear that the device can detect the 0.1 cm change in the 25 cm mixture. The 0.1 cm step is equivalent to +/- 0.4% change in the oil concentration in the mixture. Therefore, it is possible to deduce that the device has a resolution of at least +/- 0.4%. A third experiment was conducted and aimed also to test the accuracy of the device within another range of oil and water mixture. The range taken was from 20 cm water and 5 cm oil mixture to 19 cm water and 6 cm oil mixture in increments of 0.1 cm. The results are shown in Fig. 6. In the fourth experiment, another mixture range was taken Figure 2. The piezo-resistive pressure sensor and its amplification circuit Figure 3. Experimental set up 66 Vol. 3, No. 1 (2006) 63-68ring Researchrnal of EngineeThe Jou from 5 cm water and 20 cm oil stand to 3 cm water and 22 cm oil stand in increments of 0.1 cm. The obtained results are shown in Fig. 7. The experiment was conducted using four trials in order to determine the standard error and precision of the sys- tem. As an example, Table 1 displays the measurement results of the first experiment alongwith the average and standard error. The overall average precision on the pres- sure measurement was determined to be approximately 0.083 mAtm. 5. Analysis of the Experimental Results Based on the experiments conducted, it was concluded that the device can be used and trained to detect water-cut in an oil and water mixture. The results of the first exper- iment showed a clear change in the output voltage as the percentage of the oil increased in the mixture. The other three experiments showed that the device can detect small changes at different oil level ranges. In all the ranges con- sidered, the device was able to clearly detect 0.1 cm Pr es su re ( at m ) Oil level (cm) 0.028 0.027 0.026 0.025 0.024 0.023 0.022 0.021 0.020 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0.0256 0.02555 0.0255 0.02545 0.0254 0.02535 0.0253 5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5 Oil level (cm) Pr es su re ( at m ) 67 Vol. 3, No. 1 (2006) 63-68ring Researchrnal of EngineeThe Jou change of oil in the mixture. The 0.1 cm change in a 25 cm level is equivalent to a measurement resolution of +/- 0.4%. One other very important observation is that the device can operate over a full range of water-cut from 0- 100%. 6. Conclusions A number of water-cut measurement devices were eval- uated, which included capacitance and conductance devices. Other devices such as microwave, radioactive and fiber optics sensor were either unsafe or expensive cannot be used in the whole water-cut range. The pro- posed technique, which is based on the relationship between the water-cut in an oil and water mixture and the pressure of a sample from the mixture, is to be very cheap compared to the existing devices. This device is robust, user friendly and can operate in the whole water-cut range. Pr es su re ( at m ) 0.02685 0.0268 0.02675 0.0267 0.02665 0.0266 0.02655 0.0265 0.02645 0.0264 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.0222 0.02215 0.0221 0.02205 0.022 0.02195 0.0219 0.02185 0.0218 0.02175 0.0217 0.02165 Pr es su re ( at m ) 20 20.2 20.4 20.6 20.8 21 21.2 21.4 21.6 21.8 22 86 Vol. 3, No. 1 (2006) 63-68ring Researchrnal of EngineeThe Jou Experimental results showed that the new, inexpensive, and simple measuring device for water-cut in oil can attain a very high resolution that can reach up +/- 0.4% and it can be used to measure a full range of water-cut levels (0- 100%). Acknowledgements This work was supported by Petroleum Development Oman (PDO). References Beckwith, T.G., 1981, Mechanical Measurements, Addison Wesley Publishing Co. pp. 445-477. Betta, G. and D'Apuzzoand, L.M., 1993, "An Intrinsic Optical Fiber Sensor", Imeko TC4, pp. 113-119. Grattan, K.T.V., 1987, "Recent Advances in Fiber Optic Sensors," Measurement, Vol. 5, pp. 122-134. Giallorenzi, T.G., 1986, "Optical Fiber Sensors Challenge the Competition", IEEE Spectrum. Lucas, G.P., 1994, "Flow Rate Measurement in Vertical Oil-water Flows using Conductivity Sensors and a Void Fraction Wave Model," Advances in Sensors for Fluid Flow Measurement, pp. 13/1-13/3. Smit, Q., Mortimer, B.J.P. and Tapson, J., 1998, "General Purpose Self-tuning Capacitance Sensor [for Oil Recycling and Soil Moisture Measurement Application]," Instrumentation and Measurement Technology Conference, IMTC/98, Vol. 2, pp. 1074- 78. Shaofan, W.Q.D., 1994, "A High-accuracy Microwave Sensor and Calibration for Measuring Three-phase Saturations in Cores," Instrumentation and Measurement Technology Conference, IMTC/94, Vol. 3, pp. 1273-76. Shaofan, W.Q.D., 1994, "A Microwave Technique for Measuring Three-phase Saturations in Cylindrical Cores," Precision Electromagnetic Measurements, pp. 71-71. Pressure (mAtm) Oil (cm) Trial 1 Trial 2 Trial 3 Trial 4 Average Std Error 0 25.258 25.156 25.342 25.529 25.321 0.079 1 25.202 25.034 25.286 25.408 25.233 0.079 2 25.137 24.932 25.184 25.305 25.140 0.078 3 25.034 24.857 25.109 25.230 25.058 0.078 4 24.904 24.773 25.025 25.146 24.962 0.080 5 24.773 24.717 24.969 25.090 24.887 0.087 6 24.661 24.577 24.829 24.950 24.754 0.084 7 24.559 24.493 24.745 24.866 24.666 0.085 8 24.428 24.428 24.680 24.708 24.561 0.077 9 24.372 24.344 24.596 24.642 24.489 0.076 10 24.204 24.288 24.540 24.540 24.393 0.087 11 24.139 24.148 24.400 24.428 24.279 0.078 12 23.989 24.064 24.316 24.288 24.164 0.081 13 23.859 23.905 24.157 24.111 24.008 0.074 14 23.775 23.840 24.092 23.989 23.924 0.072 15 23.672 23.737 23.989 23.887 23.821 0.072 16 23.532 23.625 23.877 23.812 23.712 0.080 17 23.430 23.467 23.719 23.728 23.586 0.080 18 23.327 23.402 23.653 23.672 23.514 0.087 19 23.196 23.299 23.551 23.532 23.395 0.088 20 23.075 23.187 23.439 23.448 23.287 0.093 21 22.972 23.047 23.299 23.290 23.152 0.084 22 22.851 22.870 23.122 23.224 23.017 0.093 23 22.730 22.748 23.000 23.122 22.900 0.096 24 22.608 22.646 22.898 23.010 22.791 0.097 25 22.487 22.571 22.823 22.870 22.688 0.094 Average 0.083 Table 1. Measurement results of the first experiment along with the average and standard error values