Quick Index:
Abstract
The present study introduces an
innovative method called "Electronic Descaling (ED)," which was
developed as a means to prevent precipitation fouling in a heat
exchanger. The operating principle of the ED technology is presented
using fundamental physics laws. The validity of the ED technology was
tested by evaluating the performance of a full-scale chiller over the
entire cooling season. The ED technology eliminated scales in the
condenser tubes, resulting in the improvement of the chiller
performance and in energy savings.
Introduction
When hard water is heated in heat transfer equipment, scaling occurs. When scales deposit on a heat exchanger surface, it is traditionally called "fouling". The type of scale differs from industry to industry, depending on the mineral content of the available water. One of the most common forms of scale is calcium carbonate, CaCO3, which is the subject of the present study.
Once scales build up on a heat transfer surface, at least two problems associated with scales occur. The first problem is the degradation in the performance of the heat transfer equipment. Due to the small thermal conductivity of scales, a thin coating of scales on the heat transfer surface will greatly reduce the overall heat transfer performance. The second problem is that a small change in tube diameter substantially decreases the flow rate or increases the pressure drop across the heat transfer equipment.
Various scale-inhibiting
chemicals such as dispersing or chelating agents are used to prevent
scales [1]. Ion exchange and reverse osmosis are also used to
reduce water hardness, alkalinity, and silica level [1].
However, these methods are expensive at the industrial level, and
require heavy maintenance for proper operation. Once fouling occurs
in a heat exchanger, scales are removed by using acid chemicals,
which shorten the life of heat exchanger tubes, thus necessitating
premature replacement. When acid cleaning is not desirable, scraping,
hydro-blasting, sand blasting, metal or nylon brushing is used,
operations which incur downtime and repair costs.
Application of ED Technology to Chillers
Energy requirements for the operation of a chiller account for a significant amount of total energy requirements in building management. An effective water management program is essential in controlling energy usage. However, even a good water management program does not totally eliminate scaling. Minimum levels of even soft scale accumulation on condenser tubes will greatly reduce the chiller performance and subsequently increase the annual energy expense.
A major cause in the decrease of chiller performance during operation is fouling of condenser tubes. While manufacturers generally indicate chiller efficiency in kW/ton for new clean tubes, these values are not achieved in actual operation because of heat exchanger fouling. This is caused by the deposition of calcium and calcium bicarbonate ions on the condenser tube wall, providing a thin insulating layer, which inhibits heat transfer from the refrigerant to the cooling water.
The origin of the fouling problem lies in the hardness of cooling tower water. Even though soft water is used for make-up water, the cooling tower water quickly becomes hard due to evaporation. When the cycles of concentration in a cooling tower system are maintained at three or four, the electric conductivity of the recirculating water increases to or beyond 1,200 mS/cm, making water hard, thus creating precipitation fouling in the condenser tubes. Note that 1 m S/cm is equal to 1 micromho/cm.
Anticipating the fouling problem in the operation of a chiller, most manufacturers apply a fouling factor in a range of 0.00025 - 0.0005 hr oF ft2/Btu (1 hr oF ft2/Btu = 0.176 m2 K/W) in the design of the chiller. In an ideal chiller plant with a perfect water treatment and a perfect off-season maintenance program, it may be possible to maintain the manufacturer's design fouling factor during the entire cooling season. However, under a typical water treatment program with normal off-season brush cleaning, the actual fouling factor may be much greater than the fouling factor specified by the manufacturer. Our recent scale removal studies conducted at our laboratory showed that both nylon and brass brushes do not remove scale deposits from condenser tubes at all, once they become hard [2]. Accordingly, the energy cost for the operation of a chiller may be much greater than what has been designed, if nylon or brass brush punching is used as the normal off-season cleaning. In many plants and institutions, the cost of chiller operation is not separated from the entire electrical bill, so it is not easy to monitor the actual cost of running the chiller.
A method called "Electronic
Descaling (ED)" has been developed at our laboratory. This innovative
ED solution will improve chiller performance by preventing new scales
in the condenser tubes, translating into substantial energy savings.
Furthermore, the ED technology could remove existing scales from
condenser tubes, thus correcting the problems caused by condenser
fouling. The objective of the present study was to conduct fouling
experiments to examine whether or not an electronic descaling
technology indeed prevents fouling in a chiller, thus significantly
improving the performance of the chiller.
The Operating Principle of Electronic Descaling (ED) Technology
Figure 1 shows a schematic diagram of the operation of an electronic descaling (ED) device. A 14 gauge wire is wrapped 50 times around a feed pipe to a heat exchanger, forming a solenoid. The two ends of the wire are connected to the ED control unit. The ED unit produces a pulsing current to create time-varying magnetic fields inside the pipe. Subsequently, the time-varying magnetic field creates an induced electric field inside the pipe, a phenomenon which can be described by Faraday's law [3]:
Fig. 1 Sketch of the operation principle of electronic descaling technology
In order to maximize the induction, a pulsing current having a square-wave signal is used. The current and frequency of the square-wave signal used in the present study were 5.0 Amp and 500 Hz, respectively. More detailed descriptions on the operating principle and applications of the electronic descaling device can be found elsewhere [4-9].
The induced electric field, which oscillates with time, provides molecular agitation to charged mineral ions such that dissolved mineral ions such as calcium and bicarbonate collide and precipitate with the help of impurities in water (e.g., iron oxide particles) [10].
Ca+2 + 2HCO3- -----> CaCO3 + H2CO3
As fluid temperature increases inside a heat exchanger, precipitated CaCO3 particles and heat exchanger surface compete for dissolved mineral ions inside the heat exchanger. Since the combined surface area of the CaCO3 particles can be greater than the surface area of the heat exchanger, the fouling at the heat exchanger surface can be mitigated or prevented.
The concept of introducing fine particles to reduce fouling is not new. Troup and Richardson [11], who reviewed both chemical and physical methods of water treatment used to prevent fouling in heat exchangers, reported that fouling could be reduced by simply feeding fine CaCO3 particles or forming Al(OH)3 particles in solution using aluminum electrodes across which an alternating current was passed.
Note that the precipitation of
CaCO3 particles by the electronic descaling technology occurs in
water. Cooling tower water has excess dissolved mineral ions well
above the saturation limit of each dissolved ion, and hence the water
becomes unstable [12]. It is this supersaturated and unstable
nature of hard water that causes fouling in a heat exchanger in a
time period of months. Of course, fouling also occurs with saturated
water when it becomes supersaturated inside a heat exchanger. In this
case, the fouling takes place slowly, i.e., over a period of years,
and the use of either chemical or physical scale prevention method
can not be justified for economical reasons.
Experimental Setup
Figure 2 shows a schematic diagram of the chilled water system at the Korman Building of Drexel University. Two 450-ton centrifugal chillers (York model YT -J1K1E2-CRG) using R-123 were installed in 1996. During the 1997 cooling season, only one chiller was running at any given time, and the chillers were switched every two weeks. The total hours of operation of each chiller in 1997 was 1,500 hours. The efficiency of the YT type York chiller is 0.57 kW/ton at full-load and 0.52 kW/ton at 50-80% load [13].
Fig. 2 Schematic diagram of chiller and cooling tower arrangement. An electronic descaling unit was installed on the feed pipe into the second chiller at the Korman Building.
Blowdown was carried out using a solenoid valve which was controlled by an electric conductivity meter. The solenoid valve was designed to open when the conductivity meter hit 1,500 mS/cm, and to close when the meter read 1,200 mS/cm. Since the electric conductivity of tap water available at Drexel University was 450 mS/cm (approximately 150 ppm hardness), the cycle of concentration at the cooling tower at Korman Building was 3.0 (1,350/450).
Figure 2 also shows the application of an electronic descaling unit, which was installed on a feed pipe into Chiller #2. The ED unit treats water going into the condenser tubes of the chiller at the feed pipe non-invasively.
An automated data acquisition control system was installed in each chiller so that one can download the chiller performance data into a computer every hour. Outside air temperature, chiller motor amp, supply and return temperatures for both chilled water and condenser water were recorded from chillers #1 and #2 and later transferred to Excel files for data reduction.
In order to ensure the accuracy of temperature measurements, all thermistor probes used were calibrated. Each probe was removed from the chiller, immersed in cold water whose temperature ranged from 40 to 60oF, and the temperature reading was recorded. A calorimeter thermometer was also immersed in the same cold water bath and used as the reference for the thermistor probe calibration. Actual temperature readings were corrected based on the calibration. The error in temperature measurement was ±0.2oF.
During the entire cooling season of 1997, chillers were operated using a conventional water treatment program. The chillers were turned on May 1, 1997, and the electronic descaling device was turned on July 1, 1997. Hence, the chillers were operated without the electronic descaling device for the first two months but with the conventional water treatment program.
Method to Determine kW/ton for Chiller Efficiency
The efficiency of a chiller is usually given in terms of kW/ton, which is the electrical consumption required to generate one refrigeration ton. The former is estimated from a motor ampere measured from the chiller using the following equation:
kW = Motor Amp [A] x Voltage [V] x 1.73 x Power Factor x
where Motor Amp was read directly from the chillers used in the present study, Voltage was 480 V for the chillers, 1.73 is the constant applied to three phase AC, and the power factor was 0.93.
The unit "tons of refrigeration" is often used to represent the cooling capacity of a refrigeration system. In other words, the "ton" means the rate of heat removal from a refrigerated space. The capacity of a refrigeration system that can freeze 1 ton (907.2 kg or 2,000 lbm) of liquid water at 0oC (32oF) into ice at 0oC in 24 hours is said to be 1 ton. One ton of refrigeration is equivalent to 211 kJ/min or 12,002 Btu/hr [14].
In a normal operation of a chiller, the chilled water split, DTchilled , i.e., the temperature difference between chilled water return and chilled water supply, is measured and displayed on a chiller display panel. Hence, if one knows the flow rate of the chilled water, one can estimate the value of "ton" using the following formula:
where 8.33 is to convert gallons
to pounds, and 12,000 is to convert Btu/hr to ton. Once we know the
values of kW and ton, one can calculate kW/ton. Note that the chilled
water was circulated in the primary loop at Korman, thus maintaining
a constant flow rate during the entire cooling season, and air
handlers were used only in the secondary loop. In other words, the
flow rate in the above equation was constant throughout the cooling
season at the Korman building.
Results
Figure 3 shows the efficiency of chiller #2 before and after the electronic descaling treatment in terms kilowatts per ton at various loads. A dashed line in Fig. 3 indicates the design specification, provided for comparison. An upward arrow "A" indicates the efficiency degradation in chiller #2 during the first two months prior to the installation of the electronic descaling device. On the average, the kW/ton increased from 0.55 to 0.75, a 36% increase in energy consumption from the design kW/ton. This efficiency degradation is attributed to mineral fouling in the condenser tubes, which was happening despite an effective water treatment program.
A downward arrow "B" indicates the improvement in the efficiency of the chiller after the ED treatment started on July 1, 1997. As soon as the ED treatment started, the value of kW/ton began to drop, reaching 0.58 kW/ton on July 12, 1997. During the next month, the values of kW/ton slightly increased to 0.65 kW/ton as indicated by an upward arrow "C", and stayed at that value for the rest of the season. Since there was no forecast of the level of scaling without the ED treatment during entire season, the decrease in scaling is understated.
Fig. 3 Variation of chiller efficiency (measured in kW/ton) vs. percentage load before and after electronic descaling treatment. ED treatment began on July 1, 1997.
Figure 4 shows a typical variation of the efficiency of the chiller #2 with time before and after the electronic descaling treatment . The plot was produced using the chiller efficiency data collected in a bin of outside air temperature of 79-81oF. Note that the chiller #2 was used every two weeks for two weeks over the entire season. Day 125 represents May 5, 1997 and day 270 represents September 27, 1997. The chiller operated for 50 days (from 125 to 175) without ED treatment. During this period, the value of kW/ton increased from 0.57 kW/ton at the beginning of the cooling season to 0.82 kW/ton on June 30, 1997, as shown in closed square symbols.
When the ED treatment was installed and turned on (day 176), the performance of the chiller immediately began to dramatically improve, reaching 0.57 kW/ton within two weeks as manifested by open circles. During the next one month, the value of kW/ton gradually increased from 0.57 to 0.63 kW/ton and stayed there for the rest of the cooling season. The design specification is also shown in a dashed line for comparison.
Fig. 4 Changes in chiller efficiency (measured in kW/ton) with time before and after electronic descaling treatment. ED treatment began on July 1, 1997 (i.e., 176 days).
Figure 5 shows the condenser split, i.e., the temperature difference between outlet and inlet of the condenser as a function of % load. The condenser split curve began to shift to the left immediately after the electronic descaling treatment began. For example, at a load of 63%, the condenser split was 5.3oF on June 24, 1997. Within two weeks of the ED treatment, the condenser split at 63% load increased to 6.7oF, indicating that as the condenser scales were removed by the ED treatment. During the rest of the cooling season, the condenser splits remained unchanged.
Fig. 5 Condenser split (i.e.,
temperature difference between outlet and inlet of condenser) vs. %
load over entire cooling season with and without ED
treatment.
The present study examined the performance of an ED device in controlling the fouling problem in a chiller used in the HVAC industry. The present study clearly demonstrated that under a normal industrial condition, the chiller efficiency could decrease by 10-30% from the design specification due to condenser tube fouling. The electronic descaling device could significantly prevent the performance loss, helping chiller users save substantial energy. In the case being presented, the cost of operation of the chillers in the Korman Building would have exceeded $45,000. The actual savings realized using the electronic descaling device was $14,200, a 32% reduction.
One of the questions regarding the data presented in the present paper is the mechanism of scale prevention and removal. Since detailed descriptions of the scale prevention mechanism are given earlier, only the mechanism of scale removal will be briefly described below: As demonstrated in Fig. 4, the chiller performance dramatically improved once the electronic descaling treatment started, indicating the removal of existing scales. The ED treatment continuously produces insoluble mineral crystals, thus reducing the supersaturation level of recirculating water. As the supersaturation level decreases, the old scale layer from condenser tubes become dislodged in the form of large pieces of several inches, rendering scale-free tubes. Hence, immediately after the ED treatment, the value of kW/ton reaches the peak efficiency observed at the beginning of the cooling season. After this brief period, a thin coating of scale is deposited and removed continuously with the help of the electronic descaling treatment. Thus, the value of kW/ton increases slightly from the peak efficiency observed at the beginning of the season.
Due to the space limitation, the
performance of the chiller #1 could not be discussed in detail. The
chiller #1 was operated under almost identical conditions in the same
recirculating loop but not treated by the electronic descaling device
(see a schematic diagram shown in Fig. 2). Since the ED treatment was
at the feed pipe to chiller #2, the chiller #1 did not get the ED
treatment. However, the recirculating water was treated for two weeks
when the chiller #2 was running. Hence, there was some residual
effect of the ED treatment in chiller #1, but the benefit was
marginal.
The present study investigated the performance of an electronic descaling device in controlling fouling in a condenser used in a HVAC chiller. Two 450 ton chiller performance data were obtained through an Andover data acquisition system over the entire cooling season with and without the electronic descaling unit.
Based on the actual performance
data, it is concluded that the ED device helped a chiller to run more
efficiently (i.e., about 20-30%) even with a normal water treatment
program. The ED device not only controlled new scale build-up but
removed old scales from the condenser tubes.
