EFP Series Example Applications
Problem 1
Details
A Pressurized Water Reactor (PWR) nuclear facility has a heat exchanger under varying pressures of up to 2000 PSI (136 atm) and temperatures up to 600 °F (316 °C). Real-time measurement of void fraction / coolant density and an accurate estimate of collapsed liquid level were desired to improve reactor thermal efficiency and reactor safety.
Multiple liquid level measurement techniques had been tried, including microwave frequency guided radars, but the medium was too lossy for adequate propagation and the materials and cabling were not adequate for long term installation at these temperature and pressure ranges. Perhaps more importantly, other techniques were not able to determine coolant void fraction or estimate collapsed liquid level.
As a test of the system, a transient loss of pressure simulation was performed, and the data is shown below with explanatory captions.
Solution
With a non-conductive coolant medium, a single channel EFP unit mated to a standard Inconel® liquid level probe could have been used to measure froth height, froth-liquid interface, and adequately estimate coolant density. However, this particular medium was conductive and radar signal attenuation was too high to allow an adequate fluid density measurement from a coaxial liquid level probe. Froth height and froth-liquid interface was clearly measurable, however.
EFP Series void fraction sensors and signals are much less susceptible to signal loss in conductive media, and a vertical 4-channel void fraction sensor array was therefore chosen to determine void fraction and fluid density at multiple vertical levels. A single liquid level channel was installed.
In this situation, the coaxial liquid level probe would be used to determine froth-liquid interface, and the vertical void fraction probe array would be used to determine fluid density at multiple levels. Collapsed liquid level can be easily determined from this data.
The probes were constructed from Zirconium alloy with Inconel®-jacketed cable for the maximum signal integrity, corrosion resistance, and long life. A 5-channel EFP Series signal processor was used to read all measurements simultaneously and estimate instantaneous collapsed liquid level. The appropriate signal processing software modules were activated on each channel through the network interface. The following data was obtained:
A: Inital state. Heat exchanger is pressurized.
B: Loss of pressure transient, stage 1. Vigorous coolant boiling occurs within seconds. Measured liquid level rises dramatically (swell height), but collapsed liquid level (dashed red line) estimated from the void fraction gradient stays unchanged, confirming no loss of coolant has occurred. Pressure leak is stopped at this point.
C: Loss of pressure transient, stage 2. As the pressure is restored, the boiling begins to subside. This is reflected both in the decreased swell height and the decreased void fraction gradient.
D: Loss of pressure transient, stage 3. The fluid column has returned to near-normal following restoration of coolant pressure.
Real-time graph of data from the pressure transient experiment diagrammed above, with each letter corresponding to one of the graphical representations above. Again, notice that the estimated collapsed liquid level (dashed red) closely follows the known collapsed liquid level (dashed black). Had a loss of coolant failure occured, however, this estimated curve would have shown the loss and been the only initial indicator of loss-of-coolant failure. No other sensor system can give you this type of data.