Construction is nearly complete on one of the world's more unusual precision measurement facilities: A 50 meter horizontal smokestack located on NIST's Gaithersburg, MD campus.
Once it is fully instrumented later this year, the new facility will enable researchers to determine the most accurate methods for monitoring the amount of gases discharged from smokestacks at coal-fired power plants and other industrial sites.
At present, for a typical coal-fired power plant in the United States, measurements of carbon dioxide (CO2) flux in smokestacks have uncertainties that may be as large as 10% to 20% -- leaving actual emissions totals up in the air.
A tour of NIST's newly built scale-model smokestack simulator by mechanical engineer Aaron Johnson. The facility will help researchers understand the fluid mechanics of flows through coal-fired power plant smokestacks to accurately measure emisions.
"For both current mitigation efforts and future emissions monitoring, the nation needs to improve CO2 measurement in smokestacks of coal-burning power plants," says Aaron Johnson of PML's Fluid Metrology Group.
"And if you improve measurement of total flow, you will improve measurements of other pollutants as well, such as sulfur compounds and nitrous oxide. So we're constructing a scale-model smokestack simulator here at NIST to better understand and monitor gas flow. The goal of the project is to find a way to get CO2 flux measurements with a 1% uncertainty at a reasonable cost."
High-accuracy monitoring became even more important on June 2, 2014, when the Environmental Protection Agency announced draft plans to reduce greenhouse gas emissions from power plants by as much as 30% by 2030.
That's a titanic volume of gas to deal with. Collectively, electricity generation accounts for about 38 % of total U.S. CO2 emissions, greater than any other economic sector. There are more than 500 coal-fired power plants in the United States, ranging in capacity from a few megawatts (MW) to 2,000 MW or more. According to the Environmental Protection Agency, coal-fired plants' average emission rates are 2,249 lbs/MW hour of carbon dioxide, 13 lbs/MWh of sulfur dioxide, and 6 lbs/MWh of nitrogen oxides.
"At present, power plants are required to report their total emissions," says James Whetstone, head of NIST's Greenhouse Gas Measurements program, which is funding the SMSS project. "But in the absence of a requirement to estimate the uncertainty of the emission measurements, it is very difficult to quantitatively assess the quality of the reported data."
Determining exactly what the flow rate is within an individual smokestack (some are 10 meters in diameter), is "a tough measurement," Johnson says, both because of present metering technologies and because of an imperfect understanding of how gas velocity can vary by position in the stack.
Measurements are now made by using a Continuous Emission Monitoring System (CEMS) in each smokestack. About two-thirds of these units are ultrasonic systems made up of two transmitter/receiver units placed at different elevations in the stack. A sound signal traveling from one to the other is speeded up if it travels in the direction of the gas flow, or slowed down if it travels against the flow. The difference in transit times across that single path is used to calculate the velocity in the stack; multiplying that by the area yields flow volume.
However, the flow in plant stacks is turbulent and inhomogeneous – caused in part by the fact that the gases usually take a sharp 90-degree turn as they travel out of the scrubber and into the smokestack. "All the flow is not moving in the axial direction straight up the stack," Johnson says. "Swirling flow contributes a horizontal component to the gas velocity, and thus introduces error into velocity measurements. In general, asymmetric flows and rotating flows are bad for accurate measurement."
It appears likely that placing one or more additional pairs of ultrasonic units that sample different paths could produce more accurate measurements. So one of the goals of the new Scale-Model Smokestack Simulator (SMSS) project, which includes a section fitted with multiple separate ultrasonic transmitter/receiver units, is "to understand how much bang for the buck you get by using multi-path metering," Johnson says.
A related goal is to improve calibration of existing measurement tools. EPA requires that the CEMS apparatus be calibrated periodically according to a protocol based on velocity measurements made by two pitot-tube probes* placed facing into the gas flow stream through access ports on the side of the smokestack.
"The trouble is, we don't really know how to put uncertainties onto those measurements," Johnson says. "There are different sorts of pitot tubes, differences in precision between practitioners, and no real traceability back through every step to the internationally agreed (SI) unit of flow, which is cubic meters per second.
"We need to study different kinds of pitot tube designs and sizes and placement. Introducing a pitot tube affects the velocity you're measuring and the smaller the flow cross-section, the larger the effect would be. These blockage effects are generally negligible during use in a large smokestack, but are introduced when pitot tubes are calibrated in situ in relatively small-diameter apparatus.
"At present, those effects are poorly understood. So in the SMSS, we'll experiment with different sizes until we're satisfied that we're not changing the flow pattern that we're trying to measure. We need to assess the EPA method against reliable flow standards, as well as examine the performance of CEMS flow meters in distorted gas-flow profiles."
All this and more will take place in the new SMSS facility (see video above), which may appear quite different from a power-plant smokestack. Instead of a vertical stack, the SMSS is a horizontal arrangement. Instead of gases from combustion, the SMSS monitors air. Instead of a stack diameter around 10 meters, the SMSS pipes are about 1 meter in diameter.
Buoyancy effects are negligible: Most modern coal-fired power plants are equipped with wet scrubbers to remove sulfur oxides. After passing through the scrubber, the flue gas exhaust temperature is not substantially above the ambient temperature. Therefore, large fans force the exhaust gas up the stack
Air, a reliable proxy for flue gas, is drawn into the SMSS at controlled speed by large fans and passes through two connected sections of pipe, each about 18 meters long and set at right angles to one another to echo the path that gases in commercial plants travel from the scrubber to the base of the smokestack.
In the intake (reference) section, air enters via a bell-shaped port and travels through a 0.9 m diameter channel which includes an 8-path ultrasonic flow meter that is traceable to NIST primary flow measurement standards and has an expanded uncertainty of 0.5 % at a 95 % confidence level.
Air speed in the reference section will range from 11 to 44 meters per second with near-ideal flow conditions -- that is, a symmetric velocity profile with no swirl.
The moving air then takes a 90-degree turn and enters the 1.2 m diameter test section, where speed will vary from 6 to 25 m/s, a range common in power-plant smokestacks. Different "de-conditioning plates" will be inserted into the pipe to create controllably different asymmetric flow profiles simulating a variety of smokestack configurations. From there, the air moves into a special test segment, where the pipe can be fitted with pitot tubes and contains a multi-path ultrasonic flow test fixture.
Data collected in this facility will also be compared to computational fluid dynamics (CFD) results to evaluate the accuracy and uncertainty of this type of computer analysis. "Comparison of the CFD results to the results measured in the smokestack simulator will allow us to scale up from our four-foot diameter test section to the thirty-foot diameter section of real smokestacks" says Jacob Ricker of PML's Thermodynamic Metrology Group. "This method allows us to validate our computer models, and then use that to predict measurement uncertainties of a real smokestack without building a full sized model."
Not surprisingly, the project requires precision measurements of various critical dimensions. So the Dimensional Metrology Group (DMG) in PML's Semiconductor and Dimensional Metrology Division is working with Johnson to design a procedure for measuring transducer locations and orientation inside of smoke stacks.
The DMG has broad experience in measuring high-value large parts using laser-tracker coordinate-measuring machines and in developing uncertainty statements for specific measurement tasks. Vincent Lee and Christopher Blackburn of the DMG completed a preliminary measurement of a pipe section, including the pipe's diameter and the relative positions and locations of two ultrasonic transducers used to measure flow.
In addition to measuring the effects of different flow profiles and sensor configurations and providing benchmark data for computational fluid dynamics, Johnson says, "the SMSS will also allow us to look at alternative flow measurement techniques, putting tracers in flow, using long-wavelength acoustics, and using LIDAR."
"In order to understand the effects of pollution or to control it, you have to be able to measure it," says John Wright, a Project Leader in the Fluid Metrology Group. "The pollution control and monitoring programs established in the United States are a model for developing economies, and our stack simulator has attracted attention from researchers from other countries, including China. We are planning a workshop to coordinate our work. Our goal is to improve stack flow measurements and put solid uncertainty analyses in place for them. We need reliable stack flow data to make science-based decisions."
-- Video by Jennifer Lauren Lee
* A pitot tube is a pressure device used to measure velocity of flow in fluids. In its simplest form, it consists of a tube with an intake aperture at one end that is smaller than the diameter of the tube. A diaphragm inside the tube is deflected in proportion to the difference in pressure between the fluid inside the tube and the fluid outside.