Estimating hovering of a mobile sensor in combustion boiler environment

Measurement 109 (2017) 100–104

Contents lists available at ScienceDirect

Measurement journal homepage: www.elsevier.com/locate/measurement

Estimating hovering of a mobile sensor in combustion boiler environment I. Korhonen ⇑, V. Laine School of Engineering Science, Lappeenranta University of Technology, Opiskelijantie 3, 78210 Varkaus, Finland

a r t i c l e

i n f o

Article history: Received 7 February 2017 Received in revised form 12 May 2017 Accepted 17 May 2017 Available online 24 May 2017 Index terms: Fluidized bed Combustion modelling Mobile sensor Harsh environment Hovering sensors in fluidized beds

a b s t r a c t Our research group is investigating possibilities of developing a mobile sensor for short-term measurement work inside large industrial boilers. The aim of the sensor is to aid researchers modelling and simulating combustion processes in big boilers to test and verify their models. To that action a short-term, active mobile sensor is an extra-ordinary and unique tool, because it offers measurements in locations which cannot be reached with other traditional measurement methods in large boilers. There are many technical challenges to overcome before the active mobile sensor is ready for service. One is to know the balance between gravity forces, buoyancy and drag of the fluids from the bottom to upper parts of the boiler, in order to get the sensor ball floating and moving freely with the flows. This paper concentrates on estimating the conditions for hovering of the sensor ball. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Industrial energy and recovery boilers are ever increasing in size. The largest of them can produce up to 1500–1800 MW of thermal energy. Large energy capacities demand vast surfaces to collect the heat flow. Consequently the boilers are very big in size. The bottom area of boilers can be hundreds of square meters and the height of the combustion chamber can reach 50–90 m. [1–5]. The largest kraft recovery boiler has a bottom area of 480 m2 [1]. In order to measure inside the large boilers, new measurement methods are needed. Methods used today are unable to get information from the inner parts of the combustion chamber. In addition, new requirements for pollution control and monitoring demand new measurement methods [6,7]. Ordinary pressure, temperature and flow sensors measure only at areas near the walls. When they are traversed some meters towards the middle part of the chamber, water cooled probes must be used. These are cumbersome and uneconomic. The probe structures are heavy and demand much space outside the boiler during assembly and operation. These problems, among others, in traditional measurement techniques have led to the idea of utilizing an active, mobile sensor propagating in combustion chamber. Fig. 1 shows the opened sensor ball prototype opened developed in our research project. ⇑ Corresponding author. E-mail address: [email protected] (I. Korhonen). http://dx.doi.org/10.1016/j.measurement.2017.05.052 0263-2241/Ó 2017 Elsevier Ltd. All rights reserved.

The sensor ball contains simple electronics. The processor board is ARDUINO NanoTM board based on ARM ATmega328 processor and Arduino compatible radio module HC-12 operating at 433 MHz frequency. In addition, there is MAX6675 electronic board for connecting K-type thermo-element for external temperature measurements. Internal temperature measurements are done by PT-1000 thermistor, type DM-507. The ball has no active-cooling mechanism, but it is protected by the thermally insulating enclosure. The ball diameter was about 95 mm. The wall thickness was about 20 mm and weight varied from about 65 g to 150 g depending the enclosure material. According to first tests, the lifetime of the sensor or the time before the sensor became damaged, varied from 4 to 10 min, depending on the enclosure material. During the operational period, the sensors were able to do a series of measurements and transmit the results in real time to the base station outside the combustion chamber. In this paper we concentrate on the hovering issues of the ball. If the ball hovers in the combustion chamber, it can move to almost anywhere measuring and transmitting information from positions not reached before for measurements. The paper is arranged as follows. First boiler types and conditions inside boilers are briefly introduced. Next the basic forces affecting the sensor ball in flows and fluids are studied. Thirdly, the calculations for floating in different conditions are presented. And lastly, some conclusions are drawn.

I. Korhonen, V. Laine / Measurement 109 (2017) 100–104

101

Fig. 1. The opened mobile sensor prototype for combustion measurements. The two halves of the enclosure are fabricated from Skamol-1100E insulator board.

2. Conditions inside the boiler furnaces The combustion chamber is the volume in boiler work, in which the chemical energy of fuels is converted to thermal energy. The energy is collected through water and steam cooled walls and transported in steam form to turbines and heat exchangers. The turbines generate electricity. The rest of the heat from the heat exchangers can be delivered to industrial processes and district heating. There are a variety of burning processes and boiler types. Main types of boilers are circulating fluidized bed (CFB), show in Fig. 2, bubbling fluidized bed (BFB), grate boilers and kraft recovery boilers (KRB). While all are used for energy production, the KRB is used also for chemical recovery from black liquor [4]. In BFB and CFB boilers there is a large quantity, up to tens of tons, of sand inside the boiler, which is used as fluidized bed material. In bubbling fluidized bed sand fluidizes and form bubbles that are about one to two meters in diameter over the gas distributor grid. In the CFB boiler, the sand is fluidized and it circulates from

Fig. 2. Cross section of a CFB boiler.

grid up to upper parts of the boiler and through cyclone and back-loop back to the grid. In both boiler types, sand evens out the burning process and temperatures. CFB boilers operate at a relatively higher riser gas velocity. In the CFB boiler, the sand transports the heat energy to heat exchange surfaces. The floating conditions in CFB boilers differ a lot from the conditions in other boiler types [2]. The boilers differ from each other in suspension densities inside the combustion chamber. Some researchers have classified circulating fluidized beds into low density (LDCFB) and high density (HDCFB) boilers [8]. In LDCFB boilers, at bottom level of the bed the sand suspension densities are hundreds of kilograms/m3 [2]. The density decreases gradually upwards in the region of bed and continues to decrease in riser as a function of distance above the grid to 10–20 kg/m3, see Fig. 3 [9]. Near the walls, extending to some tens of centimeters from the walls, there are sand streams flowing down [2]. These must be taken into account in practical hovering situations. In BFB boilers, the bed density is significantly greater than in CFB boilers, up to 1000–2500 kg/m3 [2]. Above the bubbling bed, in the splash zone area, the density decreases precipitously, see Fig. 4. In the upper parts of the combustion chamber, suspension

Fig. 3. Suspension density as a function of distance from grid in CFB boiler.

102

I. Korhonen, V. Laine / Measurement 109 (2017) 100–104

Fig. 5. Cross-section of a power plant. Temperatures in the boiler area can rise up to + 1400 C, which means lower air and gas densities [1,3,4].

Fig. 4. The density profile of BFB boiler shows very low densities over char bed and splash zone.

density reaches a stabilized level. In that level the density is typically defined by temperature dependent density of flue gas and particles. Density is only about tens of grams more than flue gas density at operational temperatures in those parts of the boiler. Studies has shown that the suspension density is about 400– 500 g/m3 [2]. In kraft recovery boilers, at the bottom a large heap of unburned material, often called char bed, deposits in the grid. It contains mainly carbonaceous char, sodium sulfate and sodium carbonate, small amounts of potassium and trace elements. The density of the heap varies from about 2000 kg/m3 (smelt) to as low as 400– 1300 kg/m3 at top layer of the heap. The char bed is typically 1– 2 m in height [4]. In the upper parts of the kraft recovery boiler the suspension density is defined by black liquor spraying or droplets, unburned particles and gases. The particle and ash concentration in the flue gas is about 10–25 g/m3 N [10], meaning that the flue gas density is about 0.4 kg/m3. There is no information available describing fluid or gas concentration in the middle parts of KRB boilers. The fluid velocities inside combustion chambers are dependent on the boiler type. In the riser area of BFB boilers superficial velocities are 0.5–2.5 m/s and in CFB boilers 4–6 m/s [2,11]. At least one source gives a gas velocity of 3–10 m/s for CFB boilers [12]. In boilers that use pneumatic transport, e.g. pulverized coal boilers (PCB), fluid gas velocities can be 15–30 m/s. [2]. In KRB boilers fluid velocities at riser area are about 3–5 m/s. In the heat exchanger areas gas velocities are much greater, typically 10 – 30 m/s, in all type of boilers [4,12]. Common to all power and recovery boilers is that inside them the conditions are very harsh. Temperatures inside the combustion chamber reach up to +1400 °C [2,4,5]. Fig. 5, gives typical temperatures of CFB and BFB boilers. The high temperatures means that gas densities inside the combustion chambers are a lot lower than in room temperatures. This in turn has an influence on gas viscosity and drag force.

3. Calculating hovering conditions When an object is freely set in a flowing medium, such as air and water, basically three types of forces act on the object. They are, if we neglect non-vertical forces, drag FD, buoyage FB and gravitation force FG, as shown in Fig. 6.

Fig. 6. Three main forces affect sensor ball (yellow) in hovering: gravitational, drag and buoyancy forces. Non-vertical forces are neglected.

The object will hover in the upward direction, if there is a the balance between forces acting in the up and down directions, or

FG ¼ Fd þ FB

ð1Þ

Gravitation force is defined as a product of mass m of the object and earth’s gravity g or

F G ¼ mg

ð2Þ

The object experiences lift or buoyance force according to Archimedes law, which states, that the object which is partially or completely immersed in a fluid is buoyed up by a force FB which equals the weight of the fluid the object is displacing. So the buoyancy is

F B ¼ V qg

ð3Þ

where V = volume of the object immersed in the fluid, q = density of the fluid. If the fluid around an object is in motion, the viscosity of the fluid, which is analogous to friction between solid objects, causes a drag force FD to the object. This drag force is calculated by different ways depending on the fluid and turbulence. In that sense fluids are classified according to dimensionless Reynolds number, which is defined as a ratio between inertia forces and viscous force of a fluid, in form of equation:

Re ¼

quL l

where

ð4Þ

103

I. Korhonen, V. Laine / Measurement 109 (2017) 100–104

q = density of the fluid,u = characteristic velocity between fluid and object, L = characteristic linear length, m = viscosity of the fluid. If the Reynolds number is less than 2000, the flow of the fluid is laminar. If the Reynolds number is greater than 4000, the flow is turbulent. Between these limits is the inter-mediate region, where the flow can be turbulent or non-turbulent. The drag force can be calculated for a spherical object in a non-turbulent fluid using to Stokes equation F D ¼ 6pr lu

Table 2 Parameters for equation [2]. Range of Re

Region

a1

b1

0 < Re < 0.4 0.4 < Re < 500 500 < Re

Stokes law Intermediate law Newtons law

24 10 0.43

1.0 0.5 0.0

ð5Þ

wherer = radius of the (spherical) object, m = viscosity of the fluid,u = velocity. In turbulent flow conditions the drag force is calculated using the density of the fluid, viscosity and drag-coefficient CD, as;

F D ¼ CD

Aqu2 2

ð6Þ

In Eq. (6) A is the surface area of the object seen from the direction of flow or projected area. For a spherical object

A¼

pd2

ð7Þ

4

where d = diameter of the sphere. Drag coefficient is a dimensionless number which has been defined by empirical tests for different types of objects. Typical values for drag co-efficient are given in the Table 1 [13]. The drag coefficient for sand particles in CFB boilers is given by Basu [2] as a function of Reynold’s number by Eq. (8).

CD ¼

a1

ð8Þ

Reb1

The constants a1 and b1 can be approximated according to the Table 2 [2]. As we see, in the Newtons law’s region, substituting parameters in the Eq. (8), we get CD = 0.43 which is close to the value given for a spherical object in the Table 1. 4. Estimating hovering The estimations were started by modelling and simulating velocity and pressure fields related to a ball hovering in flowing air in a round channel, which diameter was 500 mm. The diameter of ball was set to 95 mm. Simulation work was done with ComsolTM simulation tool. In simulations the fluid temperature was set to +850 °C, and air viscosity and density were set as the values at that temperature i.e. 4.1 ⁄ 105 Pas and 0.309 kg/m3, respectively. The velocity field in the middle axis of the channel is shown in Fig. 7. The simulation results show that in this round channel the ball of diameter 500 mm has no impact to the flows near the channel walls and vice versa at velocity of 6 m/s and less. The tail of disturbance effects lasts more than 1 meter on the leaving side of the ball.

Table 1 Examples of drag coefficients [13]. Type of object

CD

Sphere Cube, against surface Cube, against corner Equilateral triangle, 2D Cone, top to flow, 30o Disc, flow to normal of surface

0.47 1.07 0.81 1.6 0.55 1.17

Fig. 7. Velocity field around the sensor ball at air flow speed of 6 m/s.

For estimations the hovering drag forces the ball experiences in the hovering situation were calculated using a similar sensor ball as in our first lifetime tests, see Fig. 1. The sensor ball diameter is 95 mm. The mass of ball varies depending on the enclosure wall density. For calculations the mass of the ball was set to 60 g. In floating tests only the mass and size (diameter) of the sensor ball are known. Roughness of the ball surface was neglected in this study, because of its’ relative small effect at steady state region of drag coefficient [14]. In the estimations the ball is kept static at a position and rising flows from bottom to top of a combustion chamber are used. The ball properties are fixed, and only flow speeds and fluid or suspension densities are varied in different regions of different types of boilers based on the information introduced earlier in this paper. Fig. 8 shows the results for a CFB boiler. Drag forces for a ball, whose mass is 56 g and diameter 95 mm, in air at NTP and CFB boiler were calculated with fluid densities of 1.273 (kg/m3), 5 kg/m3 and 10 kg/m3. The calculated corresponding buoyage forces are only 0.00285 N, 0.0119 N and 0.0223 N, respectively. Density values 5 and 10 kg/10m3 exist inside the chamber except in the bed, where the densities are more than a decade higher. According to Fig. 8, at fluid densities of greater than 10 kg/m3 the ball can hover when fluid velocity is 6 m/s. Velocities like this exist mainly in fast fluidized bed reactors. In typical fluid velocities present in CFB boilers, the mass of ball must be smaller or fluid densities higher. Of course, the ball will float on the bed in every situation, because the suspension density of bed can be hundreds of kilograms/m3 [2].

104

I. Korhonen, V. Laine / Measurement 109 (2017) 100–104

addition, short review of fluid densities and fluid velocities in different kinds of boilers was presented. The clear result is, that the sensor ball like used in evaluation, does not hover anywhere else except in fast velocity CFB or HDCFB boilers. The sensor floats on beds of all boilers. The results of the study are valid only to the prototype of the sensor ball, whose mass was 56 g and had a diameter 95 mm used here. In the next phase, mass of both electronics and enclosure protecting it, must be reduced. For electronics it means using one printed board and a lighter battery. For the enclosure the mass can be to some extent reduced by making the enclosure walls thinner or by selecting new materials. References

Fig. 8. Simulated drag forces vs. fluid velocity in a CFB boiler. 1. Ball gravity force, 2. Fluid density 1273 (kg/m3) = air, 3. Air-sand suspension, fluid density 5 kg/m3, 4. Air-sand suspension, fluid density 10 kg/m3.

In BFB boilers the ball can hover only on the top of the bed and just over it, in the splash area, where the suspension density vary between bed suspension density values, hundreds of kilograms/ m3, to free board area suspension densities of 400–500 g/m3 [2]. Elsewhere in BFB boiler both fluid densities and fluid velocities are too low for hovering. In kraft recovery boilers the ball doesn’t hover. On the top of the char bed the ball will adhere to the char bed material. So this sensor ball could be the first measuring device capable of measuring parameters on the char bed surface! 5. Conclusions In the paper hovering of a new sensor ball intended to operate inside big combustion chambers was studied. The sensor ball is designed to measure temperatures, pressures, flows, ionization and some other parameter defined later in the project. The main types of forces affecting the hovering sensor ball were studied. In

[1] Andritz, Andritz Showcase: Kraft Recovery Boilers, OKI Pulp @ Paper Mills, Indonesia, 2015. [2] P. Basu, in: Combustion and Gasification in Fluidized Beds, Taylor & Francis Group, 2006, p. 473, ISBN 978-0-8493-3396-5. [3] A. Hotta, K. Kauppinen, A. Kettunen, Towards new milestones in CFB Boiler Technology- CFB 800 MWe/New 460 MWe Super-Critical Plant with CFB Boiler in Lagisza - First Experience Update. Foster Wheeler Energia OY, Finland, 2010, p. 20. [4] E. Vakkilainen Kraft recovery boilers- principles and practice. Picaset Oy ISBN 952-91-8603-7 2007 Helsinki 246 [5] S. Teir: Steam Boiler Technology. Helsinki University of Technology. ISBN 9512267594. 2003, p. 216. [6] Environmental Protection Agency. Standards of Performance for Greenhouse Gas Emissions from New, Modified, and Reconstructed Stationary Sources: Electric Utility Generation Units. Final rule. RIN 2060-AQ91. Federal Register, USA, 2015. [7] World Energy Outlook 2015 Factsheet. Global Energy Trends to 2040. International Energy Agency (IEA), 2015, p. 1. [8] A. Issangya, D. Bai, H. BI, K. Lim, J. Zhu, J. Grate, Suspension densities in highdensity circulating fluidized bed riser, Chem. Eng. Sci. 54 (1999), 5461 5460. [9] P. Mirek, J. Ziaja, The influence of sampling point on solids suspension density applied in scaling of hydrodynamics of a supercritical CFB boiler, Chem. Proc. Eng. 32 (4) (2011) 391–399. [10] P. Mikkanen, Fly Ash Particle Formation in Kraft Recovery Boilers. VTT Publications 421, Finland, 2000, p. 71. [11] V. Tanskanen, CFD Study of Penetration and Mixing of Fuel in a Circulating Fluidized Bed Master Thesis, Lappeenranta University of Technology, Finland, 2005, p. 175. [12] S. Teir, A. Jokivuori, Thermal Design of Heat Exchangers. Energy Engineering and Environmental Protection Publications. Helsinki University of Technology. Finland, 2000, p. 24. [13] F.M. While, Fluid Mechanics, WCB/McGraw-Hill, Singapore, 1999, p. 826. [14] J.M. Cimbala, Drag of Spheres. Penn State University, USA, 2012, p. 3. Printed 9.5.2017 .