Fluenta (Norway)

We present measurement results of capacitive micromachined ultrasonic transducers (CMUTs) in permanent contact mode over a wide pressure range (1 - 8 atm). The CMUT plates are in contact with the bottom of the cavities due to atmospheric pressure, even without any dc bias voltage. The electrical input impedance at various dc bias voltages are measured at elevated pressure to characterize individual devices. The open and short circuit resonant frequencies are extracted from the impedance data, and the acoustic performance of pairs of devices is evaluated by performing pitch-catch measurements. A frequency matching method is proposed and used to determine the optimal dc bias voltages for the transmitting and receiving CMUTs individually. Our electrical impedance results show good agreement with the finite element model results (modal and harmonic analysis performed with ANSYS) over the entire pressure range. Moreover, the pitch-catch measurement results validate the proposed frequency matching method for an optimal biasing scheme, and a received signal with good signal-to-noise ratio of 45 dB was observed at a pressure of 7 atm. In conclusion, the behavior of CMUTs in permanent contact mode can be predicted well with our FEA, and they are indeed a promising solution in providing ultrasonic transducers that can operate over a wide pressure range.

SAND MONITOR OVERVIEW The subsea ultrasonic sand monitor described usesa clamp-on transducer that measures the acoustic energy generated by sand grain impacts on the inner side of a pipe wall (See Figure 1). A preamplifier is connected to the acoustic transducer by a miniature coaxial cable. The preamplifier is connected to a signalprocessing board that outputs a 4-20 mA current signal in logarithmic proportion to themeasured acoustic energy. The components are packaged in a sub sea housing (Figure 3) that is attached by means ofmounting adapters clamped to the outside of the pipeline. The detector signal output is carried to the sub sea electronics module via a 2-pair oil filled cable before telemetry to surface via the umbilical. The topside master control station combines the acoustic energy signal with flow velocity information an outputs the mass of produced sand. SANDMONITOR MEASUREMENT PRINCIPLE Although the transducer is especially sensitive to acoustic emissions due to particle impact, it also reacts to flow induced noise (background noise) as shown in Figure 11. Since noise related to flow velocity is an unwanted component of the signal, the sandmonitor filters out this noise with an interference canceling function, subtracting it from themeasured signal. The remaining signal represents that generated by sand. So far, ithas proven difficult to reconcile theory with measurement for the multitude of different flow conditions encountered. Consequently, a background noise canceling function must be empirically derived by fitting a polynomial curve to observed detector output signal levels recorded at approximately 5 different flow rates (Shown in Figure 14). The flow velocity is calculated in the topside master control stationusing a model that relates choke position, well head temperature and pressure to flow rate. The detector signal output generated per unitof sand production, mA per gram/see. has been determined to be unique within reasonableaccuracy. The signal corresponding to a particular sand production rate is affected by the flow velocity. For this reason the system uses a normalized sand signal vs. flow velocity function to convert sand generated noise to sand production rate given that flow velocity is known. The normalized sand signal vs. velocity function is determined by subtracting the background noise from the total output signal observed while a known amount of sand hasbeen injected into a test flow loop. The total detector signal while circulating 5 grams of sand in the flow loop is shown as a function of velocity in Figure 12. The resulting normalized sand noise function is shown in Figure 15. To give an example of the described operation, lets imagine that the detector outputs a value of 8 mA to the topside master control station, while flow velocity is 3 mtrs./sec. At this velocity the master control station will subtract4.25 mA (read out from Figure 14) from the detector signal. The remaining 3.75 mA is divided by the normalized sand signal at 3 mtrs./sec. which is 0.25 mA per. gram/see(read off from Figure 15).