CEL3 -Double Pipe Heat Exchanger PDF

Preview

Summary

Lab Report for Double Pipe Heat Exchanger...

Description

Double Pipe Heat Exchanger NAME: Hashim SURNAME: Najem MATRIC NUMBER: s1330198

DATE: 1

Table of content Abstract ................................................................................................................................................... 3 Introduction ............................................................................................................................................ 3 Theory ..................................................................................................................................................... 5 Experimental ........................................................................................................................................... 7 Results ................................................................................................................................................... 10 Discussion.............................................................................................................................................. 11 Conclusions ........................................................................................................................................... 13 References ............................................................................................................................................ 13 Appendix A: Sample Calculations.......................................................................................................... 14 Appendix B: Error Analysis .................................................................................................................... 15 Appendix C: Data Sheet ........................................................................................................................ 17

2

Abstract The objective of this experiment was to see how the heat transfer coefficient of water varies with changes to design, liquid flow rates and operating conditions. It was found that liquid flow rate, and hence the turbulence of the flow, had the largest impact on the heat transfer coefficient, as seen by heat exchanger 4 which had the highest flow rate in the shell side and the highest outer heat transfer coefficient, ho, was the highest, with ho being 4700±612 W m-2K-1. It was also found that baffles were the most effective heat exchanger design modification as they reduced sagging in the inner tube, increasing the inner heat transfer coefficient, hi. This is shown by the baffled heat exchanger, heat exchanger 3, having the highest hi, with a value of 9000±1390 W m-2K-1.

Introduction Heat exchangers are ubiquitous. They can be found in your home in air conditioning and boilers, and in every industrial process you can think of. They come in all shapes and sizes, and there are too many different designs to list. A common heat exchanger in industry is the shell and tube heat exchanger. This is a heat exchanger where there is an outer tube which has one fluid passing through it and one or more inner tubes with another fluid passing through them. Even these heat exchangers have many variations.

Figure 1: Example of a baffled straight-tube heat exchanger [1]

The performance of a heat exchanger is based on multiple reasons such as, but not limited to: • the materials of construction, • velocity of the fluid flow, • whether the fluid is turbulent or not, • the fluids in question, • whether the heat exchanger is baffled or not. Heat exchanger performance will change with the baffle pitch, which is the spacing between the baffles but for this experiment, this variable is not tested. [2] Fouling also affects heat transfer resistance but in this experiment, it is neglected.

3

The effects of these variables can be summed up by the overall heat transfer coefficient, U. From the overall heat transfer coefficient, the individual heat transfer coefficient for the inner and the outer fluid in that specific heat exchanger, hi and ho respectively, can be found. The relationship between the overall heat transfer coefficient and the individual heat transfer coefficients is covered in more detail in the Theory section. The objective of this experiment was to see how the heat transfer coefficient of water varies with changes to design, liquid flow rates and operating conditions. This is achieved by seeing how 4 different heat exchangers perform when the flow rate of cold and the hot water, flowing in the shell side and tube side respective, is varied. The effect of counter current, when the hot and the cold streams enter at opposite end, as opposed to co-current flow, when the hot and cold stream enter at the same end. The inlet and outlet temperatures for the fluid will be measured and the flow rates will be noted and this data will be used to find the overall heat coefficient and the Reynolds’ number, which is a ratio of momentum to viscous forces of the fluid flow, to find the individual heat transfer coefficients. Many papers have been publish exploring how heat transfer coefficients vary, an examples of one is the paper by E. Akpabio, I.O. Oboh, E.O. Aluyor, "The Effect of Baffles in Shell and Tube Heat Exchangers", 2009. They specifically focus on the effect baffles have on the heat transfer coefficient.

4

Theory A heat exchanger is ‘a device which transfers heat from one medium to another’ [3], in this case from hot water to cold water. It functions based on the first law of thermodynamics which states that energy cannot be created or destroyed, only converted from to another or transferred from one location to another. We can calculate the energy transferred using the following equation (McCabe et al, 2005): 𝑄 = 𝑐𝑝 𝑚󰇗∆𝑇

[1]

Where Q is the rate of heat transfer in W, cp is the specific heat capacity of the substance in J K-1 kg, m is the mass flow rate in kg s-1 and ΔT is the temperature change of the substance in K. Heat exchangers are made from different materials and have different designs. The heat transfer resistance associated with a heat exchanger can be found using the following equation(McCabe et al, 2005): 1

𝑄 = 𝑈𝐴∆𝑇𝑓

[2]

Where U is the overall heat transfer coefficient in W m-2K-1, whose reciprocal is the heat transfer resistance. A is the area of heat transfer in m2 and ΔTf is the log mean temperature difference which can be found by the equation (McCabe et al, 2005): ∆𝑇𝑓 =

𝜃1 −𝜃2 𝜃 ln 1

[3]

𝜃2

Where θ1 is the temperature difference at one end of the heat exchanger and θ2 is the temperature difference at the other end, both in K. The heat transfer resistance is due to 3 things, with the absence of fouling factor: resistance of film of the inner fluid, resistance of inner tube wall and the resistance of the outer fluid. This phenomenon can be represented by the following equation (McCabe et al, 2005): 1 𝑈

1

𝑑

1

=ℎ + 𝜆+ℎ 𝑖

[4]

𝑜

Where hi is the inner fluid heat transfer coefficient in W m-2K-1, d is the thickness of the inner tube in m, λ is the heat conductivity of the material of the inner tube in W m-1K-1 and ho is the outer fluid heat transfer coefficient. Any heat transfer coefficient can be found empirically using the DittusBoulter correlation for calculating the Nusselt number (McCabe et al, 2005): 𝑁𝑢 =

ℎ𝑑 𝜆

1

= 0.023𝑅𝑒 0.8𝑃𝑟 3

[5]

Where h is the heat transfer coefficient in question, Re is the Reynolds number and Pr is the Prandtl number, d is the characteristic length, in this case it is the diameter of the tube in m and λ is the thermal conductivity of the fluid in W m-1K-1. The Reynolds number and Prandtl number are dimensionless ratios described by the following equations (McCabe et al, 2005): 𝑅𝑒 =

𝑣𝑖𝑠𝑐𝑖𝑜𝑢𝑠 𝑑𝑖𝑓𝑓𝑢𝑠𝑖𝑣𝑖𝑡𝑦 𝑢𝜌𝑑 𝑚𝑜𝑚𝑒𝑛𝑡𝑢𝑚 , Pr = = 𝜇 𝑣𝑖𝑠𝑐𝑖𝑜𝑢𝑠 𝑓𝑜𝑟𝑐𝑒𝑠 𝑡ℎ𝑒𝑟𝑚𝑎𝑙 𝑑𝑖𝑓𝑓𝑢𝑠𝑖𝑣𝑖𝑡𝑦

Where u is the fluid velocity in m s-1, μ is viscosity in kg s-1 m-1, ρ is density in kg m-3, d is the characteristic dimension, in this case diameter in m.

5

Equation [5] can be rearranged to form: 1

ℎ

=

1 𝑑 𝑃𝑟 3 𝑅𝑒 0.8 0.023𝜆

= 𝑎. 𝑅𝑒 −0.8

[6]

Everything except the Reynolds number is a constant therefore can be simplified to a single coefficient, a. If equation 6 is combined with equation 4 and one of the fluid flow rates is kept constant, it can be shown as: 1 = 𝑏 + 𝑎. 𝑅𝑒 −0.8 [7] 𝑈 Where b is: 1 𝑑 𝑏= [8] + ℎ𝑐𝑜𝑛𝑠𝑡

𝜆

With hconst being the heat transfer coefficient of the constant flow fluid. Equation [7] can be seen to be a linear graph if 1/U and Re-0.8 are the vertical and horizontal axis respectively and the intercept is b which can be used to find the heat transfer coefficient of the constant flow fluid.

6

Experimental

7

Apparatus

Figure 2: Schematic of Apparatus [4]

The configuration of the heat exchangers shown in figure 1 is given below: • HE1: Cross-flow heat exchanger • HE2: Wired mesh heat exchanger (shell side, 20 coils/turns) • HE3: Baffled heat exchanger (shell side) • HE4: Stainless steel inner tube heat exchanger • HE5: Plain copper heat exchanger The cold water is just mains water and the hot water is generated using stream at 2 barg at the heat exchanger at the top. The steam flow rate can be controlled by valve V2 and the pressure using valve V3. The inlet flowrate of cold water is controlled using valve V4 and passes through a rotameter to tell its flow rate before entering the heat exchangers in the shell side. The inlet for the hot water is controlled by valve V6 before passing through another rotameter and entering the heat exchangers in the tube side. V1 is used to switch between co-current and counter current flow in the shell side. The inlet and outlet temperatures for shell and tube side are measured using thermal couples and the temperatures can be read using a computer. The sixth heat exchanger at the bottom is used to lower the temperature of the hot water before it reaches the collection tank at the bottom. The water is recirculated around the system by a cold and a hot water pump, which can be turned on using switches B1 and B2 respectively.

Experimental Procedure Before the experiment was started, the dimensions of the heat exchangers were measured using copies presented to us with digital callipers, which had an error of ±0.005mm. Valves W5 and K5 was opened and all other W and K valves were closed. This meant HE5 is used as the heat exchanger for the first run. Valve V1 was adjusted to counter current flow. Before the hot water pump is turned on, the valve for the hot water by-pass, V5, must be open. It can be adjusted to further manipulate the hot water flowrate but never closed, otherwise the pump may be damaged.

8

Steam was let into the hot water heat exchanger using V2 and cold water is let in using valve V4. The water pumps, B1 and B2, should also be switched on. There are 3 tasks for fulfil for each heat exchanger: • Task 1: Constant flow of cold water, counter current 1. Set cold water flow to 20%. 2. Set hot water stream flow to 5 cm on the rotameter scale. 3. Wait 2 minutes for temperatures to stabilise. 4. Record shell and tube inlet and outlet temperatures using computer. 5. Increase hot water flowrate by 5-6 increments until 15 cm. • Task 2: Constant flow of hot water 1. Set hot water flow to 12 cm on the rotameter. 2. Set cold water stream to 20%. 3. Water 2 minutes for temperatures to stabilise. 4. Record shell and tube inlet and outlet temperatures using computer. 5. Increase cold water flowrate by 5-6 increments until 65% (maximum achievable flow). • Task 3: Constant flow of cold water, co-current 1. Adjust valve V1 to allow co-current flow 2. Set cold water flow to 20%. 3. Set hot water stream flow to 5 cm on the rotameter scale. 4. Wait 2 minutes for temperatures to stabilise. 5. Record shell and tube inlet and outlet temperatures using computer. 6. Increase hot water flowrate by 5-6 increments until 15 cm. Repeat tasks 1, 2 and 3 for heat exchangers 2, 3 and 4. Due to time constraints, only 4 heat exchangers could be tested. The error of the temperature readings was ±0.25 K because of fluctuations in the temperature even after it stabilised slightly. The error in the cold water flow was ±1% and the error in the hot water flow was ±0.2 cm due to fluctuations in both readings.

9

Results The heat transfer coefficients for the inner and outer fluids can be seen in the following graphs, they were calculated using the intercept of the overall resistance vs Reynolds’ number graphs in Appendix C, see Appendix B for sample calculation.

Heat Transfer Coefficients of Outer Fluid (h o) 6000 5000 4700 HO (W M-2K-1)

4000 4000 3500

3000

3200 2500

2000

2400

2200

1000

1100

0 HE2

HE3

HE4

Counter Current

HE5

Cocurrent

Figure 3: Heat transfer coefficient for the outer fluid film.

Heat Transfer Coefficients of Inner Fluid (h i) 45000 40000

HI (W M-2K-1)

35000

36000

30000 25000 20000 15000 10000 5000

9000

7000

7000

0 HE2

HE3

HE4

HE5

Figure 4: Heat transfer coefficient for the inner fluid film.

The inner and outer heat transfer coefficients are also showed in tabular form with errors below: ho counter current (W m-2K-1) ho co-current (W m-2K-1) hi (W m-2K-1) Heat Exchanger 2 2200±308 1100±148 7000±1220 3 3500±450 3200±516 9000±1390 4 4700±612 2500±347 36000±5830 5 4000±530 2400±391 7000±1060 Table 1: Heat transfer coefficients of inner and outer fluid film.

10

Discussion The aim of the experiment was to find the heat transfer coefficient of water as an inner and outer fluid in various heat exchanger configurations. This aim was met, values for hi and ho were obtained. See from figure 3, the heat transfer coefficient for the outer fluid, ho, was higher for the counter current than for the co-current for each heat exchanger. This correlation disagrees with theory as the heat transfer coefficient should stay the same if the only change is whether the heat exchanger is co-current or counter current. This discrepancy is not due to error either as it is consistent through the heat exchangers. The difference is also too large to be considered due to error. The counter current heat transfer coefficient for heat exchanger 2 is double the value for co-current, 2200 and 1100 W m-2K-1 respectively, while the error is only 308 and 148 W m-2K-1 respectively. A theory for this difference is because the actual velocity of the liquid in the shell varies significantly between co-current and counter current. This is because the difference in distance the water must travel for co-current as opposed to counter current. The difference in distance travelled plus any extra bends will mean that the pressure drop on the fluid due to the tubing will be different for the different configurations, hence different velocities. The theory also states that heat exchanger 3 should have the higher outer heat transfer coefficient. This is because baffles increase the velocity of the fluid and allow for more turbulence. [5] This is the case for the co-current run, with the outer heat transfer coefficient being 3200±516 W m-2K-1. For the counter current run, this is not the case with heat exchanger 3 having the second lowest outer heat transfer coefficient, 3500±450 W m-2K-1, with heat exchanger 4, the steel heat exchanger, having the highest outer heat transfer coefficient with 4700±612 W m-2K-1. A theory about this anomaly is that heat exchanger 3 has a smaller shell cross-sectional area, hence the smallest effective shell diameter. When looking at the Reynolds equation: 𝑅𝑒 =

𝑢𝜌𝑑 𝑚𝑜𝑚𝑒𝑛𝑡𝑢𝑚 = 𝜇 𝑣𝑖𝑠𝑐𝑖𝑜𝑢𝑠 𝑓𝑜𝑟𝑐𝑒𝑠

We can see the Reynolds number, hence how turbulent the flow is, is a function of fluid velocity and diameter. Fluid velocity itself is inversely proportional to the square of the diameter, because fluid velocity is the fluid flow divided by the area the flow is passed through. Although it might seem at first glance that if the diameter of the shell were to decrease the Reynolds number should decrease, since the fluid velocity increases by the square as diameter decreases, the net effect of a decrease of diameter would yield an increase in Reynolds number. Because the Reynolds number increases, hence the flow is more turbulent and the heat transfer coefficient would be higher because of this increased turbulence. See Appendix C for tube dimensions. Heat exchanger 2 does not follow the trend. It has the smallest cross sectional area, hence should have the most turbulent flow, but it has a wire coil in the shell side and this wire coil would have as a resistance to the flow, flowing the fluid velocity significantly. This would make it the least effective configuration and this is evident by looking at figure 3 as heat exchanger 2 has the lower outer heat transfer coefficient for both the co-current and counter current flow, with 1100±148 W m-2K-1 and 1100±148 W m-2K-1 respectively.

11

Looking at figure 2, the inner heat transfer coefficient, hi, for heat exchangers 2, 3 and 5 are similar with values of 7000±1220 W m-2K-1, 9000±1390 W m-2K-1 and 7000±1060 W m-2K-1. This similarity is since all their inner pipes are identical, and since they all have water passing through them at the same velocity, they should be similar. Heat exchanger 3 has a slightly higher inner heat transfer coefficient, this could be chalked off as due to the large errors involved, but this increase in heat transfer coefficient is because of the baffles. The other, and more important, function of baffles in a heat exchanger is to hold up the inner tubes and prevent sagging. They also reduce vibrations due to fluid flow, especially at high velocities. [5] The absence of baffles in heat exchanger 2 and 5 means that the sagging reduced the inner heat transfer coefficient and hence the effectiveness of the heat exchange. This could be due to the sagging slowing the fluid down, hence making it less turbulent and less effective at heat transfer. The result for the inner heat transfer coefficient for the heat exchanger 4 is erroneously large, almost 4 times as large as the other heat exchangers. This must have been an error in the experiment as in defies all logic. The magnitude of the errors in these results range between 13% to 18% with the error in the inner heat transfer coefficient being generally the highest relative error. These errors could be considered large but it was inherent in the experiment. The pressure of the steam to heat the hot water would vary between 2.5 bar to 3 bar and hence the temperature of the hot water would vary significantly. This means the uncertainly in the temperatures was large, ±0.25 K per reading, with temperatures fluctuating constantly even after the heat exchanger was left to stabilise for 2 minutes. This fluctuation in temperature would compound the error as temperature readings were used in both equations [1] and [2].

12

Conclusions The aim of this experiment was to find out how the heat transfer coefficient of water in the outer tube and the inner tube would vary varies with changes to design, liquid flow rates and operating conditions. It was found that the outer heat transfer coefficient would vary between counter and co-current, this was because the actual fluid velocity though the shell would not be the same for counter and co-current because the different paths the fluid passed through for counter and co-current. Maximising the fluid velocities in the heat exchanger would allow for the largest heat transfer coefficient, both inner a...