Validation of Electronic Anti-Fouling Technology in a Single-Tube Heat Exchanger

prepared by                                                         Under Construction

Professor Young I. Cho, Ph.D., Department of Mechanical Engineering & Mechanics

Drexel University, Philadelphia, PA 19104

URL:         http://httpsrv.ocs.drexel.edu/faculty/choyi

  ABSTRACT

The objective of the present study was to investigate the validity of an electronic anti-fouling (EAF) technology in the mitigation of precipitation fouling in a once-through flow system with a single-tube heat exchanger.  Effects of flow velocity and water hardness on the effectiveness of the EAF technology were experimentally studied.  The water hardness varied from 750 to 1,000 mg/L as CaCO₃.  For both 750 and 1,000 mg/L solutions, the EAF treatment reduced fouling resistance by 20-38 % at a flow velocity between 0.52 and 0.78 m/s.  As the velocity decreased to 0.28 m/s, the EAF treatment was not helpful in reducing fouling.  SEM photographs of scale produced from the 1,000 mg/L solution at 0.78 m/s indicated that calcium carbonate scales without the EAF treatment were needle-shaped aragonite, which is sticky, dense, and difficult to remove.  Scales with the EAF treatment showed a cluster of elliptic shape crystals.  Both the heat transfer test results and SEM photographs obtained in the present study support the validity of the EAF technology in mitigating precipitation fouling.

 Nomenclature

A = heat transfer surface area

A = cross sectional area vector 

B = magnetic field strength vector

d_t,o = outside tube diameter

D_{H  = hydraulic diameter}

D_{s = inside shell diameter}

E = induced electric field intensity vector

_{Q = heat transfer rate}

s = line vector along the circumferential direction

T_{c,in = inlet temperature of cold water}

T_{c,out = outlet  temperature of cold water}

DT_LMTD = log-mean-temperature-difference

U = overall heat transfer coefficient based on outside tube diameter

U_clean = overall heat transfer coefficient at clean state

 

Experimental Method

Figure 2 schematically shows the test facility which consists of two reservoir tanks, two pumps, two flow meters, a static mixer, an electronic anti-fouling control unit, a solenoid coil, the main heat transfer test section made of a shell-and-tube heat exchanger, and a plate-and-frame heat exchanger.  

Most previous fouling studies [8-13] were conducted using a recirculating flow system.  The critical drawback of conducting a fouling test using a recirculating system is the continuous reduction of hardness during the test.  Indeed, most fouling problems in the real world occur in a once-through flow system.  In order to create the best simulation of fouling problems occurring in the real world the present study conducted fouling test in a once-through flow system.

The main heat transfer test section is made of two concentric copper tubes, which form a counter-flow heat exchanger.  Two reducing unions were used at both the inlet and outlet of the heat exchanger for easy connection between the heat exchanger and connecting pipes.  Hard (cold) water was pumped to the shell side whereas hot water moved through the tube side.  Hence, scales were deposited on the shell side only.  The shell diameter (ID) was D_s = 0.01684 m whereas the tube diameter (OD) was d_t,o = 0.01275 m, rendering the hydraulic diameter of D_H = 0.00409 m.  The diameter of both the inlet and outlet connecting tubes was 0.0109 m.  The axial length of the tube in the heat exchanger was 0.7 m.  

Since the hardness of tap water available in Philadelphia is approximately 150 mg/L as CaCO₃, it is not suitable for fouling experiments.  Therefore, hard water in a range of 750 to 1,000 mg/L was prepared in a laboratory.  For example, in order to prepare 1,000 mg/L CaCO₃ solution, 0.01 M calcium chloride (CaCl₂) and 0.02 M sodium bicarbonate (NaHCO₃) were added to tap water.  Table 1 gives values of alkalinity, electric conductivity, and pH of each concentration solution.  The resulting hard water caused severe scaling inside a pump, and one pump was lost during almost every test due to the pump scaling.  In order to avoid the scaling in the pump, the calcium chloride and sodium bicarbonate solutions were prepared in two separate reservoir tanks, pumped using two separate pumps and two flow meters, and later mixed using a static mixer as shown in Fig. 2.    

The inlet temperature of the hard water was maintained at 302±0.5^oK for all tests.  Flow rates varied from 2.71 x10^-5 m³/s (0.43 gpm) to 7.57x10^-5 m³/s (1.2 gpm), which resulted in the shell-side velocity of 0.28 and 0.78 m/s, respectively.  The Reynolds numbers corresponding to 0.28 and 0.78 m/s were 1,870 and 5,020, respectively.  It is of note that the viscosity and density of the hard water used in the Reynolds number calculation were evaluated at an average temperature of the hard water.   

Thermocouples used in the present study were Omega model TMTSS-125G-6 (grounded copper-constantan T type).  Calibration was carried out at 273 and 373^oK, confirming the manufacturer's claim of the accuracy of ±0.1^oK.  The temperature of the hot stream entering the tube side varied from 355 to 367±1.0^oK in order to study the effect of hot water temperature on the fouling resistance and initial fouling rate.  To produce hot water continuously, a plate-and-frame heat exchanger was used, where high pressure steam provided by the Philadelphia city steam network was used.  Flow meters were also calibrated over a range of 6.3 x10^-6 to 6.3 x10^-5 m³/s by measuring the weight of water accumulated over time, and the calibration results are reported elsewhere [14].  

The overall heat transfer coefficient, U, was calculated as a function of time using the log-mean-temperature-difference (LMTD) [15], which was obtained from the four temperatures measured at both the inlet and outlet of cold and hot streams.  Fouling resistance, R_f, was calculated using the usual definition [15].       

        Q = U A DT =  C_p (T_c,out - T_c,in)                                                              (3)

 

        U =                                                                                           (4)

and

        R_f = -                                                                                                    (5)

       

where U_clean is the overall heat transfer coefficient corresponding to the clean state, i.e., at t = 0.  The U_clean values for 0.28, 0.52 and 0.78 m/s cases were 2,460, 3,430, and 4,130 W/m²K, respectively.  

The errors estimated from the uncertainty analysis for Q, U, R_f^, and A (heat transfer surface area) were 2.5%, 3.0%, 8.0%, and 1.1%, respectively.  During the course of the present fouling experiment, it was discovered that the accuracy of the outlet temperature reading could be significantly affected by the scaling of the thermocouple probe itself [16].  Therefore, the thermocouple measuring the outlet temperature of the cold water was cleaned every time before each temperature measurement.  It is speculated that much of the previous fouling data in the literature might have been seriously affected by the probe fouling because the thermocouple fouling could easily cause a temperature measurement error by several degrees.  

Results and Discussion  

Figure 3 represents variation in fouling resistance as a function of time at a flow velocity of 0.78 m/s (1.2 gpm) and a concentration of 1,000 mg/L with and without electronic anti-fouling (EAF) device.  The hot water inlet temperature was 367^oK for both cases, and the flow rate of hot water was 2.23 m/s (3.3 gpm).  The fouling resistance for both cases, as manifested by the slopes in Fig. 3, increased linearly with time, indicating that the fouling was mostly controlled by a diffusion mechanism.  

From the slope of the curve of R_f vs. time in Fig. 3, the fouling rate without the EAF device was 1.45x10^-6 m²K/J, whereas the value with the EAF device was 1.15x10^-6 m²K/J.  At the end of 86 min. of operation, the fouling resistance without the EAF device was 0.0001287 m²K/W, whereas it was 0.0001036 m²K/W with the EAF device, a 20% drop.  

It was hypothesized that the EAF treatment  produced CaCO₃ particles, reducing the mass transfer coefficient of calcium ions and subsequently the scale deposit rate.  In order to examine this hypothesis, a scanning electron microscopy (SEM) was used to visualize crystal structures of scales.  SEM photographs given in Figs. 4a and 4b were obtained from scaled tubes produced at a flow velocity of 0.78 m/s (1.2 gpm) and a concentration of 1,000 mg/L without and with EAF device, respectively.  SEM photographs taken using specimen from other scale tubes obtained under different conditions are given elsewhere [14].   

Scale specimen for SEM was prepared with a utility knife by scraping small amounts of scales from the outlet region of each scaled tube.  SEM photographs of scales produced without the EAF device (see Fig. 4a) revealed that CaCO₃ scales were needle-shaped aragonite [17], whose dimension was approximately 25 mm by 2 mm.  Aragonite is a dangerous form of calcium carbonate scale, which is crystallized at a temperature above 303^oK [17].  It is sticky, dense, and difficult to remove. 

In contrast, the scales produced with the EAF device (see Fig. 4b) depicted a very different structure from the one produced without the EAF device.  The scales produced with the EAF device were a cluster of small elliptic shape particles (e.g., 10 mm by 3 mm) with no particular orientation, which are much easier to remove than aragonite.  The SEM photographs in Fig. 4 support the hypothesis of the EAF technology, which is to convert dissolved mineral ions into crystals in a bulk solution, thus reducing both the diffusion of dissolved ions to the heat transfer surface and subsequent precipitation reaction on the heat transfer surface.  As a result, the production of aragonite type calcium carbonate is prevented.  Calcium carbonate crystals formed with the EAF device appeared to be loosely connected compared to the one formed without it.  In other words, the scales formed with the EAF device may be removed at a small flow velocity (e.g., 2 m/s), whereas the scales formed without the EAF device may not be removed even at a large flow velocity (e.g., 5 m/s).  The determination of the exact threshold flow velocity for both cases is currently under way and will be reported in the future.  

Figure 5 represents variation in fouling resistance as a function of time at a flow velocity of 0.52 m/s (0.8 gpm) and a concentration of 1,000 mg/L with and without EAF device.  The hot water inlet temperature maintained to 361^oK for both cases, and the flow rate of hot water was 0.74 m/s (1.1 gpm).  In spite of a reduced flow velocity of cold water (i.e., 0.52 m/s), the fouling resistance leveled off after approximately 90 min. of operation, reaching asymptotic fouling resistances of 0.0001287 m²K/W and 0.0001036 m²K/W for the cases without and with the EAF device, respectively.  The asymptotic fouling resistance obtained with the EAF device was 23% less than that without the EAF device.  

Figure 6 shows variation of fouling resistance as a function of operating time at a cold water flow velocity of 0.28 m/s and a concentration of 1,000 mg/L with and without EAF device. The hot water inlet temperature was 367^oK for both cases, and the flow rate of hot water was 2.23 m/s - a severe fouling condition.  At the end of 240 min. of operation, the fouling resistance obtained with the EAF device was almost identical to the one obtained without the device.  The EAF treatment at this reduced flow velocity was found to be not much useful in mitigating fouling.  

In order to understand why the EAF treatment did not reduce fouling under these severe fouling condition, one can examine the fouling curve in Fig. 6.  Initial fouling rate without the EAF device was greater than that with the EAF device (i.e., see the slopes of the fouling curves at t = 0).  Hence, more scales deposit on tube surface without the EAF device, reducing scale surface temperature.  Subsequently, less scales deposit due to the decrease in the surface temperature, as indicated by arrow “A” in Fig. 6.  On the other hand, less scales deposit initially with the EAF device.  Subsequently, the temperature drop in the heat transfer surface is less, and fouling rapidly progresses as indicated by arrow “B” in Fig. 6.  As enough scales deposit on the tube surface eventually, the surface temperature drops, leading to asymptotic fouling resistance.   

Figure 7 presents variation in fouling resistance as a function of time at a flow velocity of 0.78 m/s (1.2 gpm) and a concentration of 750 mg/L with and without EAF device.  The hot water inlet temperature was 94^oC for both cases, and the flow rate of hot water was 2.23 m/s.  At the end of 89 minutes of operation, the fouling resistance without the EAF device was 5.086x10^-5 m²K/W, whereas the value with the EAF device was 3.78x10^-5 m²K/W, a 26% drop.  Compared with the results given in Fig. 3, the percentage drop in the fouling resistance due to the EAF device was greater for the 750 mg/L solution than for the 1,000 mg/L solution, suggesting that as a fouling condition becomes less severe the benefit of the EAF treatment may become large.   

In order to verify this suggestion, the inlet temperature of hot water was reduced from 367^oK to 355^oK, and the corresponding test results obtained with and without the EAF device are shown in Fig. 8.  The flow velocities of cold and hot water were 0.65 and 2.23 m/s (1.0 and 3.3 gpm), respectively.  At the end of 105 min. of operation, the fouling resistance without the EAF device was 5.25x10^-5 m²K/W, whereas the value with the EAF device was 3.26x10^-5 m²K/W, a 38% drop.  As the hardness of cold water decreased from 1,000 to 750 mg/L, the percentage drop in the fouling resistance due to the EAF device increased from 20 to 38%.  Of note is that the hardness of cooling tower water is usually maintained at an electric conductivity between 1,200 and 1,500 mS/cm, which is equivalent to a hardness of approximately 450-500 mg/L.  Hence, the EAF treatment should reduce the fouling resistance in a heat exchanger used in any cooling tower system by at least 40%.  

Figure 8 also shows a peculiar fouling behavior during the first 20 minutes of operation for both cases with and without the EAF device.  As soon as tests started, the fouling resistance became negative for both cases, indicating that the overall heat transfer coefficient, U, increased from that of the clean state.  This peculiar phenomenon has not been uncommon [18].  This can be explained as follows:  The nascent fouling deposit produces a roughened heat transfer surface.  As the surface becomes rough, the convective heat transfer coefficient increases through better mixing and breaking the laminar sublayer at the wall.  Hence, the overall heat transfer coefficient initially improved, resulting in negative fouling resistances.  As fouling deposits continue, the fouling layer acts as a thermal insulator, decreasing the overall heat transfer coefficient.  The fouling resistance begins to increase and becomes positive at t = 13 min. without the EAF device and at t = 23 min. with the EAF device.  The trend of the fouling resistance having negative values initially must have occurred for other tests whose results are shown in Figs. 3-6.  It is speculated that this period of the heat transfer improvement was too brief to observe. 

Conclusions 

The present study was conducted in order to investigate the validity of electronic anti-fouling (EAF) technology, which was considered as a means to control new precipitation fouling in a heat exchanger.  Fouling tests were carried out in a counter-flow shell-and-tube heat exchanger, and the fouling resistances obtained without the EAF device was compared with those obtained the EAF device for two different water hardness (750 and 1,000 mg/L) and three different flow velocities (0.28, 0.52 and 0.78 m/s).   

Scanning electron microscopy (SEM) photographs of scale produced without the EAF device revealed needle-shaped aragonite (i.e., sticky, dense, and difficult to remove), whereas the scale produced with the EAF device produced a cluster of loosely-connected small elliptic shape particles (i.e., easier to remove than aragonite).  

For the 1,000 mg/L solution at a velocity of 0.78 m/s, the fouling resistance with the EAF device was 20% smaller than that without it.  As the hardness of water decreased from 1,000 to 750 mg/L, the percentage drop in the fouling resistance due to the EAF device increased from 20 to 38%. However, at a reduced flow velocity at 0.28 m/s or less, the EAF treatment was not helpful in the mitigation of fouling.  

Although the present paper uses calcium carbonate as an example of the mineral scales, the electronic anti-fouling treatment is not limited to the calcium carbonate scale.  This is because the EAF treatment utilizes the electrical charges of dissolved ions to foster collision and precipitation into insoluble crystals.  

References  

 

1.    Tchobanoglous, G. and Burton, F., Wastewater Engineering, Treatment, Disposal and Reuse, 3rd edn., McGraw-Hill, New York, 1991.  

2.    Serway, R.A., Physics for Scientists and Engineers, 3rd ed., Saunders College Publishing, Philadelphia, PA, 1990, p. 874-891.  

3.    Fan, C.F., A Study of Electronic Descaling Technology to Control Precipitation Fouling. Ph.D. thesis, Drexel University, Philadelphia, PA, 1997.  

4.    Fan, C.F. and Cho, Y. I., Microscopic Observation of Calcium Carbonate Particles: Validation of an Electronic Anti-Fouling Technology. Int. Comm. Heat Mass Transfer, 1997, 24, 757-770.

5.    Fan, C.F.  and Cho, Y. I., A New Electronic Anti-Fouling Method to Control Fouling. National Heat Transfer Conference, Vol. 12. Baltimore, Maryland, 1997, p. 183-188.  

6.    Cho, Y. I., Fan, C.F., and Choi, B.G., Theory of Electronic Anti-Fouling Technology to Control Precipitation Fouling in Heat Exchangers. Int. Comm. Heat Mass Transfer, 1997, 24, 747-756.  

7.    Cho, Y. I., Choi, B.G., and Drazner, B.J., Use of Electronic Descaling Technology to Control Precipitation Fouling in Plate-and-Frame Heat Exchanger. Compact Heat Exchangers for the Process Industries, (ed. by R.K. Shah), Begell House, New York, 1997, p. 267-273.

8.    Hasson, D., Avriel, M., Resnick, W., Rozeman, T., and Windreich, S., Mechanism of      Calcium Carbonate Scale Deposition on Heat Transfer Surfaces, Int. Eng. Chem. Fund., Vol. 7, 1968, p. 59-65.

9.    Sheikholeslami, R. and Watkinson, A.P., Scaling of Plain and Externally Finned Heat Exchanger Tubes, J. of Heat Transfer, Vol. 108, 1986, p. 147-152.

10.  Watkinson, A.P. and Martinez, O.,  Scaling of Heat Exchanger Tubes by Calcium Carbonate, J. of Heat Transfer, 1975, p. 504-508.

11.  Watkinson, A.P., Water Quality Effects on Fouling from Hard Waters, in: Heat    Exchangers - Theory and Practice, Taborek, J. et al., Hemisphere, 1983, p. 853-861.

12.  Andritsos, N., Kontopoulou, M., and Karabelas, A.J., Calcium Carbonate Deposit Formation under Isothermal Conditions, Can. J. Chem. Eng., Vol. 74, 1996, p.911-919.

13.  Rabas, T.J. and Tarborek, J., 1996, Heat-Rate Improvements Obtained by Retubing Condensers with New, Enhanced Tube Types, J. of Enhanced Heat Transfer, Vol. 3, p. 83-94.

14.  Choi, B.G., A Study of Fouling Control in Heat Exchangers with Electronic Anti-  Fouling Technology, Ph.D. thesis, Drexel University, Philadelphia, PA, 1998.

15.  Incropera, F. P. and DeWitt, D.P., Fundamentals of Heat and Mass Transfer, 4th Edition., Wiley, New York, 1996.

16.  Cho, Y.I., and Choi, B.G., Effect of Fouling on Temperature Measurements Error and a Solution, J. of Heat Transfer (in press).

17.  Cowan, J.C., Weintritt, D.J., Water-Formed Scale Deposits, Gulf Publishing Company, Houston, TX, 1976.

18.  Bansal, B., and  M ller-Steinhagen, Crystallization Fouling in Plate Heat Exchangers, J. of Heat Transfer, Vol. 115, 1993, p. 584-591.

Table 1  Values of alkalinity, electric conductivity, and pH of each concentration solution.  Also shown are the amounts of calcium chloride CaCl₂ and sodium bicarbonate NaHCO₃ to tap water for 750 and 1,000 mg/L solutions.

 

Concentration         (mg/L)	CaCl2     (mol)	NaHCO3     (mol)	Alkalinity      (mg/L)	Conductivity      (µS/cm)	pH
750	0.0075	0.015	760-810	3,060-3,190	7.88-8.10
1,000	0.01	0.02	980-1,040	3,860-3,950	7.81-8.05

 

List of Figures

Fig. 1   Schematic diagram of the concept of electronic anti-fouling (EAF) technology.  EAF control unit produces a solenoid-induced molecular agitation through Faraday’s law and is installed in a feed pipe prior to heat exchanger. 

Fig 2.   Schematic diagram of a once-through flow system with a counter-flow single-tube heat exchanger.  An EAF solenoid coil is located between pump and the inlet of the main heat transfer test section.

Fig. 3   Fouling factors vs. time for two different cases: without and with electronic anti-fouling (EAF) device.  Concentration of test solution = 1,000 ppm; cold water flow velocity = 0.78 m/s; hot water flow velocity = 2.23 m/s.

 

Fig. 4   Photographs taken using a scanning electron microscopy with scale specimen from scaled tubes produced at a flow velocity of 0.78 m/s and a concentration of 1,000 mg/L (a) without electronic anti-fouling device and (b) with EAF device.  Magnification = 1,500. 

 

Fig. 5   Fouling factors vs. time for two different cases: without and with electronic anti-fouling (EAF) device.  Concentration of test solution = 1,000 ppm; cold water flow velocity = 0.52 m/s; hot water flow velocity = 0.74 m/s.

 

Fig. 6   Fouling factors vs. time for two different cases: without and with electronic anti-fouling (EAF) device.  Concentration of test solution = 1,000 ppm; cold water flow velocity = 0.28 m/s; hot water flow velocity = 2.23 m/s.

 

Fig. 7   Fouling factors vs. time for two different cases: without and with electronic anti-fouling (EAF) device.  Concentration of test solution = 750 ppm; cold water flow velocity = 0.78 m/s; hot water flow velocity = 2.23 m/s.

 

Fig. 8   Fouling factors vs. time for two different cases: without and with electronic anti-fouling (EAF) device.  Concentration of test solution = 750 ppm; cold water flow velocity = 0.78 m/s; hot water flow velocity = 2.23 m/s.

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig. 1

 

 

 

 

 

 

 

 

 

 

 

 

Fig. 2