Electric Field Effects on Carbon Monoxide Release from Impinging Flames

Transcription

1 Paper # 070MI-0360 Topic: Microcombustion and New Combustion Devices 8 th U. S. National Combustion Meeting Organized by the Western States Section of the Combustion Institute and hosted by the University of Utah May 19-22, 2013 Electric Field Effects on Carbon Monoxide Release from Impinging Flames Chien, Y.C Dunn-Rankin, D. Department of Mechanical and Aerospace Engineering University of California, Irvine, California Carbon monoxide results from incomplete combustion. In some syngas processes the carbon monoxide is desirable but usually it is a dangerous emission from forest fires, gas heaters, gas stoves or furnaces where the insufficient oxygen available in the core reaction does not fully oxidize the fuel to carbon dioxide and water. Electrical aspects of flames, specifically, the production of chemi-ions in hydrocarbon flames have been investigated in a wide range of applications. Despite this fairly comprehensive effort to study electrical aspects of combustion, relatively little research has focused on electrical phenomena near flame extinguishment. This research examines the use of electric fields as one mechanism for controlling combustion as flames are partially extinguished when impinging on nearby surfaces. In particular, it studies the use of the electrical properties of the flame to determine the combustion behavior and it then explores the use of the electric field driven ion wind to improve the burning in real time. 1. Introduction and motivation This research studies the potential of controlling the release of carbon monoxide from impinging flames by using electric fields. Inside a hydrocarbon flame, positive ions and negative charge carriers (generally electrons) appear in the reaction zone. Proper application of electric fields can act on these charge carriers and modify combustion behavior. Electrical aspects of flames have been studied for many years, with continuing investigations in microgravity and zero gravity combustion[1] [5]. These past studies have helped demonstrate several different experimental methods for measuring the ions generated in flames and their influence[6]. Despite this fairly comprehensive effort to study electrical aspects of combustion, however, the only published study on electrical phenomenon near flame extinguishment was proposed at the 2012 International Symposium on Combustion by Weinberg et al.[7]. The paper includes measurements of carbon monoxide concentration with respect to each location of a diffusion flame to evaluate the possibility of developing a sensor for surface proximity by detecting electrical currents in the flame. The research focused on sensing and so mainly supplied the flame with low DC voltage from batteries (37 and 56 volts) and probed with an electrode around a partially premixed flame to observe the first ion current appearance. The experiment was also conducted with different electrode materials. The results showed that aluminum is chemically and catalytically active with the generation of CO near quenching while brass, copper or steel are catalytically inactive materials. The electrical structure of quenching flames is also studied under small applied (<10V/mm) electric fields, and the current detection area is compared with the quenching regime where CO release is maximum. The catalytic

2 effects are significant since they can lead to erroneous interpretations of the carbon monoxide formed by the flame directly as compared to that produced during interaction with the quenching surface. This prior research revealed the possibility of using a low voltage source with different probe materials as a CO detecting sensor, and also opened another field to investigate electric field control of CO release. To begin exploring the effects associated with impinging flames interacting with an electric field as an active control mechanism, this paper investigates high voltage electric field effects on CO release near a quenching surface. 2. Experimental The experiments are conducted using a coflow burner with a plate progressively lowered toward the burner surface so that it gradually quenches the flame. In order to generate a diffusion flame, a stainless steel coflow burner is used (Figure 1). The burner is 13 cm tall with a 4 cm outer diameter; it sits on a Teflon mount to prevent conducting current through the optical table. At the exit, the burner has a 2.13 mm inner diameter center small tube ducting fuel and, at the same time, air is provided separately through a concentric outer ring annulus. The air is designed to form a uniform distribution (top-hat) at the exit after it passes through a bed of beads and a honeycomb mesh (25 mm in inner diameter) close to the exit. Inside the burner, the length of the tube is sufficient to ensure fully developed flow at the exit for the range of desired Reynolds numbers. The quenching plate is 2 inches by 2 inches square thin stainless steel with 2 mm thickness. It is indirectly connected onto a vertically moving slide. The slide can be adjusted in the range up to 40 mm from the burner exit. An inverted L-shaped connection bridges between the plate (with insulation) and the slide. The plate is screwed through two cylindrical ceramic threaded posts for electrical insulation, which are connected onto the adjustable mount. The diffusion flame in this research is stabilized with a constant speed 20 cm/s for both methane and air at the burner exit. Fig. 1 Coflow burner with a quenching plate The carbon monoxide emission is measured with a portable combustion analyzer (TSI CA- CALC 6200, figure 2) that aspirates the exhaust gas inside the chamber with a slow and fixed flow rate. The generated CO from quenching fills the chamber and mixes with the existing air yielding the 2

3 natural release from the chamber. Once the CO concentration in the chamber no longer increases, a steady-state, fully mixed condition is confirmed. The acrylic chamber was designed to create a steady environment for the experiment and to diminish the external influence of air currents on the small flame. The entire schematic of the system and setup is shown in Figure 3. Fig. 2 Portable combustion analyzer and the emission probe inside acrylic chamber Fig. 3 The measurement system schematic diagram. (1) PC (2) High voltage power supply (TREK) (3) Acrylic chamber (4) Emission meter (5) Data acquisition (DAQ) box (6) Shunt current system (7) UCI coflow burner (8) Digital camera (9) Movable quenching plate 3

4 3. Electric Field Control over Impinging Flames High voltage is applied onto the quenching plate electrode from a high voltage power supply which ranges from -10,000 volts to +10,000 volts. A computer is connected which sends an output request to the high voltage power supply. In hydrocarbon flames, H 3 O + is the most important ion for near stoichiometric mixture combustion, having a much higher concentration than other ions such as C 2 H 3 O +, C 3 H 3 + and HCO +. Prager, et al. found H 3 O + dominant in lean flames, and this finding has been confirmed over a wide range of combustion conditions, including stoichiometric ones. The fundamentals of chemi-ion production and the transport of these ions is well-documented in the references already cited so a complete repeat of this topic is not necessary[6]. In all cases, the weak plasma naturally occurring inside flames is not essential for heat release reactions so the influence of electric fields are much less chemical than physical. The typical electric field effect on an impinging flame is shown in Figure 4 when the plate is 11 mm above the burner. At this position, the plate is very close to the flame but has not touched the outer dim luminous region. The flame is pushed down gradually as the positive charge on the plate grows (forcing a downward directed ion-driven wind). On the negative charge side, when the electric field supplies more current and ion wind, the flame is pulled up and the top of the flame narrows. The luminous zone at the flame front is brighter presumably because the ion drift wind (also called the Chattock wind) has entrained more oxidizer and shifted the reaction zone[8]. 4. Quenching & CO CO decay test Fig. 4 Electric field effect on a flame near a quenching plate at 11 mm. The carbon monoxide decay trend is examined to ensure that the concentration measured occurs at a steady condition. The test was run as the chamber filled up with CO until it was steady, matching its release from the chamber to outside. In the first four minutes, carbon monoxide concentration is recorded about every ten seconds. The resulting data has an exponential trend with time and the decay rate can be derived and curve fit with concentration as shown in Figure 5, 4

5 CO concentration (ppm) From the third minute to the forth minute, the carbon monoxide concentration drops less than ten parts per million and the decay becomes slower and flat comparing with the first three minutes. Therefore starting from the 4th minute the decay trend is steady, and this time constant is assumed for the CO measurements at different conditions y=132.5e -0.84x Time (min) data points decay trend Fig. 5 Carbon monoxide concentration decaying curve in the chamber is an exponential trend CO near impinging surface The amount of carbon monoxide is measured with the quenching plate set at different heights above the burner, as shown in Figure 6. The figure shows that when the plate progressively moves toward the burner, carbon monoxide release increases. Because of quenching, the CO concentration grows as the distance drops. Two sets of data are presented in Figure 6; the first set is measured at two minutes after the plate moves to its new position and the other set is after the CO concentration reading becomes steady (where it is not fluctuating). The time required for the reading to reach this unchanging condition differs in each case since for fixed plenum chamber volume the dilution rate will vary based on the CO release rate. Therefore, having a well-defined time scale for recording carbon monoxide concentration is necessary to ensure reproducibility in the measurement. 5

6 Fig. 6 CO concentration changing with the height of plate 5. Temperature measurement - FLIR camera Temperature and species concentration measurements in flames that have an electric field applied has been a challenge for exploring local property distribution because probes of any kind disrupt the electric field significantly. Non-intrusive measurement therefore is the only choice, and in this work, an infrared camera is surveyed to understand how the temperature distributes across the quenching plate as a function of the electric field application on the flame. The camera (FLIR SC620) measures radiative emission in the range from 7.5 to 13 microns, and it is situated directly above the quenching plate at normal incidence (to avoid complications associated with Lambert s cosine law). The camera is controlled by its native commercial software (ExaminIR, Figure 7). The quenching plate is painted with black paint and is considered a blackbody object. This assumption has been confirmed as reasonable using thermocouple temperature comparisons at a few spot locations. Fig. 7 An infrared camera connect with the software (ExaminIR) 6

7 The RGB image taken at thermal equilibrium shows a well-distributed temperature change on the plate, a typical example of which is shown in Figure 8. The temperature shown in the image does not represent a local temperature of the plate or the flame temperature. It is measuring the temperature of a thin layer of the heat resistant spray paint. The overall heat transfer includes forced convection of hot flame products on to the bottom surface of the plate, conduction through the plate, and then to the paint, the heat convection out of the surface into the air, and radiation from the surface, as the system reaches thermal equilibrium[9]. Based on straightforward heat transfer analysis it is easy to conclude that the radiation from the paint provides a reliable measure of the impinging heat from the flame on the backside of the plate. It feasible, therefore, to use the FLIR camera to characterize a quenching flame and to acquire a temperature distribution using 2-D image technology. Fig. 8 Temperature change with the flow A thermal map with no electric field applied on the impinging surface at 7 mm above the burner is shown in Figure 9 (a). The amount of carbon monoxide generates is measured as 123 ppm. Under the effect of electric power applied, a positive electric field produce from the burner to the surface, 5.7 kv/cm. The CO goes to 119 ppm with the flame lifts with ion winds entrained airs. In the other way, CO goes up to 202ppm with the flame opens wider at -5.7kV/cm. 7

8 (a) No field (b) 5.7 kv/cm 8

9 (a) -5.7 kv/cm Fig. 9 Carbon monoxide changes with thermo flux on the impinging plate as electric field applied at 7mm above the burner. This paper involves the study of how an electric field changes flame behavior near a quenching surface, and how the amount of CO release changes as the impinging surface progressively approaches the flame. In particular, the work shows how IR images of the thermal map from the impinging flame can be linked to its CO release. It also shows how the thermal map is affected by the electric field applied to the flame. This information provides evidence for the feasibility of reaching the ultimate goal of manipulating carbon monoxide over impinging flames. Remaining work requires further exploration to understand the relationship between electric field control of the flame and the heat flux effect from the surface, and ultimately how this changes the release of CO. Acknowledgements This work is supported by NASA s ISS Research Project via agreement NNX07AB55A, with Dennis Stocker as contract monitor. Disclaimer Product names and models are provided only for clarification and are in no way an endorsement on the part of NASA or the federal government. References [1] M. Papac, Electrical aspects of gaseous fuel flames for microgravity combustion, [2] S. Karnani, D. Dunn-Rankin, F. Takahashi, Z.-G. Yuan, and D. Stocker, Simulating Gravity in Microgravity Combustion Using Electric Fields, Combustion Science and Technology, vol. 184, no , pp , Oct [3] S. Karnani, S kanarni ESS - Electric field induced convection in microgravity combustion.pdf. 9

10 [4] B. A. Strayer, J. D. Posner, D. Dunn-Rankin, and F. J. Weinberg, Simulating microgravity in small diffusion flames by using electric fields to counterbalance natural convection, Proc. R. Soc. Lond. A, vol. 458, no. 2021, pp , May [5] S. Karnani, Electric field-driven flame dynamics, University of California, Irvine,, Irvine, Calif. :, [6] J. Lawton and F. J. Weinberg, Electrical Aspects of Combustion. Oxford University Press, [7] F. J. Weinberg, D. Dunn-Rankin, F. B. Carleton, S. Karnani, C. Markides, and M. Zhai, Electrical aspects of flame quenching, Proceedings of the Combustion Institute, no. 0. [8] A. P. Chattock, On the electrification needle-points of air, Philosophical Magazine, p. 285, [9] M. (Marcel) Cremers, Heat transfer of oxy-fuel flames to glass:the role of chemistry and radiation, URN:NBN:NL:UI: , Technische Universiteit Einhoven, Eindhoven,