Faster, more accurate, stack-flow measurements.

Exhaust flows from coal-fired electricity-generating plants are determined by averaging flue gas velocities measured at prescribed points in the stack cross section. These velocity measurements are made using EPA-approved differential pressure probes such as the 2-hole S-probe or the 5-hole spherical probe. Measurements using the more accurate 5-hole spherical probes require a time-consuming rotation (or nulling) of the probe to find the yaw angle. We developed a time-saving non-nulling technique using a spherical probe that measures all 3 components of velocity and therefore provides better accuracy than an S-probe. We compared the non-nulling technique with the EPA Method 2F nulling technique at both high (16 m/s) and low (7 m/s) loads in a coal-fired powerplant smokestack. Their excellent mutual agreement (within 0.3% of the flow) demonstrates that the non-nulling technique accurately measures flue gas flows.Implications: Accurate flow measurements are critical for quantifying the levels of greenhouse gases emitted from coal-fired power plant smokestacks. Flow measurement accuracy derives from the annual calibration of stack flow monitors. Calibrations are performed using EPA sanctioned pitot traverse methods called the flow relative accuracy test audit (RATA). This study demonstrates the viability of a new pitot traverse method, herein called the Non-Nulling Method. Testing in a coal-fired power plant stack showed that the new method is 5 times faster to implement than the most accurate EPA pitot traverse method (i.e., Method 2F), yet gives the same or better accuracy.

Bulk viscosity, thermoacoustic boundary layers, and adsorption near the critical point of xenon.

We present an improved model for the dissipation and dispersion in an acoustic resonator filled with xenon near its critical temperature Tc. We test the model with acoustic measurements in stirred xenon that have a temperature resolution of (T - Tc)/Tc approximately 7 x 10(-6). The model includes the frequency-dependent bulk viscosity calculated numerically from renormalization-group theory and it includes critical-point adsorption. Because the density of adsorbed xenon exceeds the critical density, the bulk viscosity's effect on surface dissipation is reduced, thereby improving the agreement between theory and experiment.

Bulk viscosity of stirred xenon near the critical point.

We deduce the thermophysical properties of near-critical xenon from measurements of the frequencies and half-widths of the acoustic resonances of xenon maintained at its critical density in centimeter-sized cavities. In the reduced temperature range 1 x 10-3<(T-Tc)/Tc<7 x 10 (-6), we measured the resonance frequency and quality factor (Q) for each of six modes spanning a factor of 27 in frequency. As Tc was approached, the frequencies decreased by a factor of 2.2 and the Q's decreased by as much as a factor of 140. Remarkably, these results are predicted (within +/-2% of the frequency and within a factor of 1.4 of Q) by a model for the resonator and a model for the frequency-dependent bulk viscosity zeta(omega) that uses no empirically determined parameters. The resonator model is based on a theory of acoustics in near-critical fluids developed by Gillis, Shinder, and Moldover [Phys. Rev. E 70, 021201 (2004)]. In addition to describing the present low-frequency data (from 120 Hz to 7.5 kHz), the model for zeta(omega) is consistent with ultrasonic (0.4--7 MHz) velocity and attenuation data from the literature. However, the model predicts a peak in the temperature dependence of the dissipation in the boundary layer that we did not detect. This suggests that the model overestimates the effect of the bulk viscosity on the thermal boundary layer. In this work, the acoustic cavities were heated from below to stir the xenon, thereby reducing the density stratification resulting from Earth's gravity. The stirring reduced the apparent equilibration time from several hours to a few minutes, and it reduced the effective temperature resolution from 60 mK to approximately 2 mK, which corresponds to (T-Tc)/Tc approximately =7 x 10(-6).

Thermoacoustic boundary layers near the liquid-vapor critical point.

We measure and calculate the sound attenuation within thermoacoustic boundary layers between solid surfaces and xenon at its critical density rhoc as the reduced temperature tau identical with (T- Tc)/Tc approaches zero. (Tc is the critical temperature.) Using the known thermophysical properties of xenon, we predict that the attenuation at the boundary first increases approximately as tau(-0.6) and then saturates when the effusivity of the xenon exceeds that of the solid. [The effusivity is epsilon identical with (rhoCPlambdaT)(1/2), where CP is the isobaric specific heat and lambdaT is the thermal conductivity.] The model correctly predicts (+/-1.0%) the quality factors Q of resonances measured in a stainless steel resonator (epsilon(ss) =6400 kg K(-1) s(-5/2)); it also predicts the observed increase of the Q, by up to a factor of 8, when the resonator is coated with a polymer (epsilon(pr) =370 kg K(-1) s(-5/2)). The test data span the frequency range 0.1