Whether working on cost reduction or product improvement, our customers often want to know how changing to our insulation will affect their product performance. At RSI we take great pride in participating in these studies. The following documentation details one of these projects.
A pool heater was supplied to us by a pool heater manufacturer to determine the suitability of our no-smoke material in the pool heater application. The heater was tested “as-is” using a special test by-pass, the designs of which were provided by the heater manufacturer. The heater was then tested using our no-smoke insulation, and the results were compared.
Our material showed a slight increased in the energy delivered to the water and a slight decrease in the “case” temperature on three of the four vertical sides of the heating unit. Based on the comparison, the RSI no-smoke material is an acceptable replacement to the refractory material currently used in the heater.
The test was designed to be performed as per the email received on August 8th, 2005 from the pool heater manufacturer. An “H” bypass was constructed from 2” type 1 PVC pipe and a 1.5 horsepower centrifugal pool pump. The H-Bypass is shown below in Figure 1.
Figure 1: H-By-Pass Test Set-Up Diagram.
Some concerns with this set-up were addressed as follows before the testing commenced.
Required pumping capacity: The unit we received requires a minimum of 40 GPM of flow (with the attendant manifold pressure) for the unit to fire. Testing the pump with the bypass closed, we achieved a flow rate of 78 GPM, midrange in the heater’s required flow parameters.
Required delivery of make-up water: With the bypass engaged, and estimating 80 GPM flow rate with 100% efficient heating of water and a 75ºF make-up water temperature (an estimate of water requirements) we needed 23 GPM of make-up water. Our water delivery capability was measured at the test site at 65 GPM.
Heater flue-gas: We do not have hood equipment ready for removal of flue gas, so the unit was placed in part of the plant with a large exhaust fan. No testing of gas was therefore possible.
Drainage capacity: The drains were tested at full delivery capacity for 10 minutes. We saw 65 GPM drain through the test room floor drains without any water back-up. This is well beyond the estimated test flow of 23 GPM.
This test was to determine the suitability of RSI insulation material in this application. As such, it was a comparative test based on the temperatures of the test unit on the outside and the vent ports around the heater, and the energy transferred to the water. We first tested the unit twice using the standard insulation, recording the inlet and outlet temperatures, and the flow rates through the test plumbing, and recording temperatures around the outside of the unit when it reached steady state (when the temperature on the topside of the heater stopped increasing). After the baseline was established, the unit was disassembled, the insulation removed, and replaced with RSI No-Smoke. The heater was re-assembled, and the tests were repeated.
Experimentation with the unit confirmed that a flow rate below 40 GPM deactivated the heater. Valves were installed in the h-bypass to act as a throttle on both the recirculation and the feed from the make-up-tank. It was found for the pump to operate without loss of flow, both the valves controlling flow into the pump needed to be wide open. This left the exhaust water valve to control the flow of heated water back into the pump. This is how the experiment was controlled.
Once the flow and recirculation rates were set, the heater was activated. The unit would only fire when the front panel was removed, and it would only stay lit for about 30 seconds at a time before a SNS (sensor error) shut down the unit. The heater would do this three times before reporting a CFH (Call for Heat) code, and not activate the firing sequence. At this time the unit was power cycled, and the errors would reset, allowing another series of three burns. This was continued until a constant temperature of 95 degrees on the inlet side could be maintained. At this point the unit would maintain fire and the front panel was re-installed. The exhaust throttle and the tank fill valve were then adjusted to make the inlet temperature 100 F and the tank level within 3 inches of the top of the tank. During 45 minutes of heating, small adjustments to the exhaust throttle were necessary to maintain the inlet temperature at 100 F. Usually, the throttle had to be opened slightly as the body of the unit heated up, and the inlet stream became hotter. This stopped after about 45 minutes, and the unit was operated for another 15 minutes to ensure that no further adjustments would be needed. At this point, flow rates and temperatures were measured on the pipes.
Unit Inlet Temperature: This temperature was reported on the control panel of the heater.
Unit Inlet Flow Rate: This information was reported from a volumetric flow meter installed downstream from the pump.
Unit Outlet Temperature: This information was taken from a handheld temperature meter.
Unit Outlet Flow Rate: This value was calculated from the mass of water filling a bucket in 10 seconds and measured on a digital scale.
Make-up Water Flow Rate: This value was taken to be the same as the outlet flow rate by conservation of mass.
Make-up Water Temperature: This information was taken from a handheld temperature checker.
After the flows and temperatures were checked, the unit itself was checked on all 5 outward exposed faces for temperature. A K-type probe pyrometer was used to check the temperature inside the vents on the sides and back, while surface contact pyrometers were used to determine the temperature on the top and front interior surface, and at surface points around the unit.
As mentioned, two issues arose during testing. We saw a SNS (sensor) error in the unit, beginning after approximately 30 seconds of firing, and repeating until the inlet stream achieved 95 degrees. Including re-start time, it took approximately 15 minutes for the unit to get to sufficient temperature to operate without the sensor error. After about one hour and fifteen minutes to an hour and a half, a HL1 error occurred in all of the tests. This is listed as a high-limit switch open error and resulted in the unit turning off. This gave about a 30-minute window to record all temperatures. The following diagram, Figure 2, shows where the temperatures were taken on the unit.
Figure 2: Temperature Recording Sites
Temperatures at locations T1, T2, T3, T4, T5 and T13 were taken with a contact type pyrometer. Temperatures at the other locations were taken with a probe-type pyrometer to facilitate getting internal temperatures under the shrouding of the heating unit. T1, T2, T3, and T4 were measured with the front exterior panel removed.
Two runs with each refractory were performed. HL1 errors prevented us from recording control temperatures during one run. A summary of the temperatures at the recording sites is given in Table 1. Control is the refractory that was shipped in the unit, while No-Smoke is RSI refractory material.
Table 1: Measurements around the Heating Unit
|Temperature Site||Measurement Type||Control 1||No-Smoke 1||No-Smoke 2|
Water flow and temperatures are shown in Table 2. There are three sets of data. During the first run of each refractory, data was logged before unit temperatures were recorded. During the second run of each refractory, flow temperatures and rates were recorded both before and after unit temperatures were taken.
|Refractory||Temp in||Temp Out||Make-up temp||Recycle rate GPM||Exit rate (lb/Hr)|
We based our conclusions on two assumptions.
1. Better refractory material will yield lower unit surface temperatures when everything is at steady state.
2. Better refractory performance will deliver more energy to the target fluid in the heat exchanger.
The first is directly measurable. The second needed a calculation to determine how many BTUs were being used to heat the water.
Comparing the unit temperature control data versus the two datasets with RSI material, we can see that the front face of the unit with RSI material tended to run slightly hotter than the control, while the rest of the unit ran a little cooler with RSI refractory. Overall the RSI material tended to run slightly cooler, although the magnitude of the temperature difference was below 1%.
To determine heating efficiency, we needed to calculate how much energy the water absorbed when the unit was running. The energy balance we need is:
Min(Tin) + Q = Mout(Tout)
Where Min is the mass flow rate into the heater in Lb/Hr, Tin is the temperature of the incoming stream in Fahrenheit, Q is the power delivered to the fluid in BTU/Hr, and Mout and Tout are the mass flow rate in Lb/Hr and Temperature (F) of the exit stream, respectively. We need to know what Q is in the above equation. Transforming the equation to solve for Q yields
Q = Mout(Tout) – Min(Tin)
In this case, the Min we can measure from the flow meter in the system (see Figure 1). By Conservation of Mass, Min = Mout. Tin can be measure from the unit controller, and Tout is measured by hand held unit. These temperatures are reported in Table 2. Converting GPM to Lb/Hr in Table 2 yields the information we need to calculate Q. Table 3 summarizes that information, along with the calculated Q.
Table 3: Energy Delivered Calculation
|Refractory||Temp in||Temp Out||Make-up temp||Recycle rate GPM||Exit rate (lb/Hr)|
The RSI material in the unit displays slightly higher energy delivery to the water being heated by about 1.5%.
The RSI material can be used as a replacement for the standard refractory based on the above data. In general there was a slight improvement in energy delivered to the heated water, and overall slightly lower temperatures recorded on the heating unit itself.