You are on page 1of 15

Chemical Engineering

School of Engineering and Physical Sciences

3rd YEAR LABORATORY FORMAL REPORT COVER SHEET


NAME: Jamie McGee

Experiment Title: Heat Exchangers

Date of Experiment: 11th February 2013

Demonstrator: Nazer Rajoub

Supervisor: Dr. G. White

For the attention of Mr Craig Bell

Please complete this form IN FULL and ATTACH your report. Failure to do so will result in it being returned to you unmarked.

Synopsis
The aim of this experiment was

Contents
Synopsis .............................................................................................................................................. 2 Theory ................................................................................................................................................. 5 Procedure............................................................................................................................................ 8 Results ............................................................................................................................................... 10 Discussion.......................................................................................................................................... 10 Conclusion ......................................................................................................................................... 13 References ........................................................................................................................................ 14 Appendices........................................................................................................................................ 15 i. ii. iii. iv. v. vi. Nomenclature ........................................................................................................................... 15 Example ..................................................................................................................................... 15 Example ..................................................................................................................................... 15 Example ..................................................................................................................................... 15 Example ..................................................................................................................................... 15 Example ..................................................................................................................................... 15

Introduction
As chemical and process engineers it is crucial that we can understand the basic concepts of heat transfer and fully appreciate how changes in the process will affect heat exchange. In industry, many chemical processes, such as reactions and separations, are generally more efficient at high temperatures. In order to raise the temperature of the feed stream(s), a heat exchanger is used to offer a medium for heat to transfer from a hot stream across a highly conductive surface to the colder stream. In this experiment three different types of heat exchangers are used. A Shell & Tube heat exchanger (two pass tube-side): Perhaps the most common type of heat exchanger in oil refineries and other large chemical processes, the shell & tube heat exchanger can be used for a huge range of applications and is operable at high pressures. Heat is transferred from one fluid to the other through the tube walls, with the tube-side fluid contacting with the tube-side fluid twice, co-currently and counter-currently.

Figure 1 http://en.wikipedia.org/wiki/Shell_and_tube_heat_exchanger

A spiral heat exchanger: In a spiral heat exchanger the fluids are contacted counter-currently and the centrifugal forces provide an increase in heat transfer capability. This type of exchanger has a number of advantages over conventional shell and tube exchangers, in additional to extending heat transfer potential, the tight configuration allows for shorter undisturbed flow. The exchanger is also easier to clean and proves significantly resistant to fouling, providing significant advantages in systems susceptible to fouling.

Figure 2 - http://www.gocesco.com/all_weldedspiralheatexchangers.html

A plated heat exchanger: The plate heat exchanger offers an increase in surface area meaning a larger contact area between the fluids is possible. This means that heat transfer occurs more quickly and more efficiently. Additionally, the plates can be removed to allow easy cleaning and maintenance of the exchanger.

Figure 3 - http://www.spiraxsarco.com/resources/steam-engineering-tutorials/steam-engineering-principles-and-heattransfer/steam-consumption-of-heat-exchangers.asp

The main objective of this experiment was to investigate the affect of varying the cooling stream on heat transfer, by calculating the overall heat transfer coefficients. These experimental heat transfer coefficients were determined by examining the temperature values at different stages in system at steady state and once calculated, allows for an efficiency to be determined by comparison with theoretical values.

Theory
At steady state, the total heat transfer and overall heat transfer coefficient can be determined from the experimental results using Equation 3. (1)
Where: Q=Heat [kJ], CP = specific heat of fluid [kJ/kg.K], m= mass flow rate [kg/s], 2 T= change in temperature (Tin Tout) [K], U = overall heat transfer coefficient [W/m K], 2 A= total heat transfer area [m ], Tm =log mean temperature difference [K], W=condensate flow rate [kg/s], hfg=latent heat of vaporisation [kJ/kg]

Equation 2 provides a method of determining U by adding the individual resistance, predicted from theory. (2)
Where: hi, ho = film heat transfer coefficients for fluids on inside and outside of the interface [W/m K], 2 k= thermal conductivity of interface material [W/m K], x= interface thickness [m]
2

Generally, the film heat transfer coefficients can be found through a variation of the correlation shown below (Equation 3). (3) Nu, Re and Pr are dimensionless groups defined by the following equations: Nusselt number: Reynolds number: Prantel number:
Where: =average velocity [m/s], =density [kg/m3], = viscosity [kg/m.s], G=mass flow rate per unit cross sectional area [kg/m.s], de=equivalent diameter [m]

Equivalent (hydraulic) diameter:

Shell & Tube Heat Exchanger


a) INSIDE TUBE hi It is assumed that only saturated vapour enters and 100% condensation takes place inside the exchanger. It is assumed that for condenser design, the film coefficient should be calculated for annular and stratified flow and the higher value selected. i) Annular flow Boyko-Kruzhilin equation [
3 2

(4) (5)

Where: L./ v = liquid / vapour density [kg/m ], kL=thermal conductivity of liquid [W/m K], 2 di=tube inner diameter [m], hi = tube side film coefficient, single-phase flow of condensate [W/m K],

ii) Stratified flow Nusselt equation *0.8

(6) (7)

Where: h= total horizontal tube loading, W= condensate flow rate [kg/s], L=tube length [m], Nt= no of tubes

b) OUTSIDE TUBE ho Kerns approach


( )

(8) (9) (10)

( )

Where: As=maximum cross flow area [m ], ds=internal shell diameter [m], lB= baffle spacing [m], C=clearance between tubes [m], Y= tube pitch [m], s= viscosity at heat transfer surface [kg/m.s],

) (11)

NEED UNITS 2. Spiral-plate heat exchanger [4] Check refs


( )

(12)

Where: W= viscosity of water at the same temperature [kg/m.s]

The viscosity correction factor (/w) in this case is 1 since water was used.

3. Gasketed-plate heat exchanger [2]

The viscosity correction factor (/w) in this case is 1 since water was used

) (13)

(14)

(15) (16) (17) (18)

(19) (20) (21) (22)

Procedure
Apparatus
The experiment is best understood when visualised. Note that the inflows and outflows from each exchanger can be controlled via a respective valve and that the temperature is taken at series of significant points along the system. APPENDIX DIAGRAM

Experimental Technique

Part A: Shell & Tube heat exchanger a) The gauge steam pressure was set to 8 psi and the cooling water flowrate was adjusted to 12 l/min, according to the rotameter calibration chart. It is important to observe these values and adjust the system to prevent any fluctuations and maintain steady state. b) The valves were adjusted to allow water to pass through the tube side and steam through the shell side of the shell & tube heat exchanger. c) Temperature values of thermocouples 1, 3 and 4 (Appendix) were recorded every minute for 20 minutes each run. d) Condensate was collected after every run for 10 seconds and the volume was measured with a measuring cylinder. e) Part A was repeated for 22 l/min, keeping all other variables constant.

Part B: Spiral heat exchanger a) The gauge steam pressure was set to 8 psi. b) The valves were adjusted to allow the cold stream (cool in) and the heated out from shell & tube (hot in) and to enter the spiral exchanger. c) The cold water inlet flow rate was adjusted to 8 l/min while the hot in stream was set at 22 l/min. d) Temperature values of thermocouples 2, 3, 4, 7 and 8 (appendix) were recorded were recorded every minute for 20 minutes each run. e) Part B was repeated varying the cold inlet flow rate to 12.5 and then 18.0 l/min, while keep all other variables constant.

Part C: Plate heat exchanger a) The gauge steam pressure was set to 8 psi. b) The valves were adjusted to allow the cold stream (cool in) and the heated out from shell & tube (hot in) and to enter the plate exchanger. c) The cold water inlet flowrate was adjusted to 8 l/min while the hot inlet was set at 22 L/min. d) Temperature values of thermocouples 2, 3, 4, 5 and 6 (Appendix) were recorded every minute for 20 minutes each run. e) Part B was repeated varying the cold inlet flow rate to 12.5 and then 18.0 l/min, while keep all other variables constant.

Results

Discussion
The aim of the experiment was to determine the efficiency of each of the respective heat exchangers and observe the influence of flowrate on the overall heat transfer coefficient, U. Upon briefly viewing the results, one can stipulate that each heat exchanger shows a unique response when comparing the analytical results to the experimental findings. Overall, efficiency tends to be low across the board, ranging from an average of 17.3% to 56.8%. In order to effectively analyse the results, each heat exchanger must be considered in its own right as the pipeline and design features vary between exchangers and the unique design aspects of each exchanger, which differentiate between one another, not only allow for different applications in industry but will also react differently to fluctuations in the environment, such as fouling and loss of heat energy in piping. The Shell & Tube heat exchanger shows an average of 56.8% across the two different flow rates. It can be seen that both the average U value and the efficiency show a direct proportionality with changes to flow rate however it is difficult to justify this finding with only two data points recorded. Ideally, a minimum of three data points should be recorded in order to verify a correlation however due to the circumstances present in the laboratory, this was not possible. A large source of source of error can be attributed to the assumption made which states that only saturated vapour enters the tubes and that all of this condenses entirely. This assumption is false and has a direct impact on the results calculated as the values extracted from the steam will be inaccurate, introducing an error carried through the calculations. As the mass flow rate of condensate and the U value are correlated in a linear fashion, for every drop of condensate that was wrongly included in the value, the U will be inaccurate by a directly proportional amount. Incorrectly induced condensation refers to steam condensing outside the exchanger in the pipelines either before or after entering the exchanger. The Boyko-Kruzhilin equation also makes this same assumption in calculating inside film heat transfer coefficient, compounding this error further. It also worth noting that while conducting the experiment there fluctuations in the steam pressure as the pressure would frequently fall from the target pressure of 8psi to as low as 6psi, even with the mains supply at the full setting, highlighting another major flaw in the design of experimental rig. This fluctuation again invalidates some of the information extracted from the steam tables, as a change in the pressure will impact the thermodynamic properties of water. As previously touched upon, this stage of the experiment would be more constructive were more readings taken and it is recommended that the technical issues surrounding the availability of the mains inlet supply is resolved should further experimentation occur. It is important to consider the importance of fouling within a heat exchange system and in this scenario no data was given regarding the cleaning or maintenance of any exchanger on the experimental rig. The theoretical analysis does not incorporate any potential for fouling and if fouling were to be present (it is likely) then an additional layer of resistance must be accounted for, and a lower U would have been determined. In this experiment the main source of fouling is likely to be due to the erosion and corrosion of the exchangers and solid deposits within the unit. When

sizing a shell & tube heat exchanger, the effect of fouling must be considered as they are particularly difficult to clean in comparison with other type. If frequent cleaning is required, then the process engineer should consider installing a unit of greater diameter, as this means it can be cleaned with greater ease. The spiral heat exchanger shows an average of 39.1% efficiency across the three data points established in the experiment. This shows a lower average efficiency value than that of the shell & tube exchanger. The accuracy of the spiral exchanger will suffer from the same flaws in experimental set-up and technique and the location of the spiral exchanger within the rig will amplify certain effects such as heat loss to the environment (between thermocouple and exchanger) as fluid must travel through a longer route, this will also cause a greater pressure drop across the exchanger altering the thermodynamic properties. There was also an increase in error due to further inaccuracies in the measuring equipment. For the shell & tube exchanger only a single rotameter was used to obtain a reading which can be converted to a flow rate via a calibration chart however both the spiral and plate exchangers employ the use of two rotameters, increasing the potential for error. In order to keep the flow rate constant, the valves must be manipulated in order to keep both of the rotameter floats at a constant position with reference to a visual marker. In theory, spiral heat exchangers should tend to build up less fouling due to higher velocities and greater turbulence in the channels and the relative ease of cleaning is useful for many an application. In practice, the exchangers used are aging and the level of maintenance is unknown meaning the spiral is likely to have similar level of fouling to that of the shell & tube, if not more. Again the effects of corrosion and erosion are likely to be a predominant factor however the narrow channels of the spiral could eventually become fouled with solid deposits without proper care. When analysing the graphs of U versus flow rate, the findings are again inconclusive as no trend is present meaning that further analysis regarding this aim should be conducted. The analysis of the plated heat exchanger deems it to be the most inefficient of all the three exchangers with an average efficient value of 17.3%. The sources of error indicated for the other two exchangers are present in the plate exchanger and in a similar fashion to the spiral exchanger, several sources of error are increasingly significant mainly, the heat losses from the increased pipe length and the use of two rotameters. The main advantage of the plate exchanger is the large area exposure however the heat transfer coefficient U represents heat transfer per unit area allowing a direct comparison between the exchangers. A direct comparison of the average U values shows the shell & tube and spiral to be 291.5 W/mK and 292.3 W/mK while that of the plate is significantly lower than the others at 1719.3 W/mK, meaning that the plate is on average 5.9 times larger than the other two. This is a huge deviation and indicates a deep fundamental flaw in the calculations used (theory). In future another approach should be used in order to achieve a more sensible set of results allowing for a realistic efficiency to be determined. Another potential factor in the deviation from theory could be due to a fault in the thermocouple set up around the plate exchanger, in particular thermocouple no. 6 (cooled out) appears to give an improbably high temperature. If this temperature were to be lower, this would increase the value of experimental U calculated bringing the efficiency into a more likely territory. As previously described, fouling is likely to play some kind of part in creating error however it is worth noting that a huge advantage of the plate exchanger is the ease at which it can be cleaned. The U versus flow rate graph shows a clear trend across three results that would point to direct proportionality between the two elements. Unfortunately, this finding loses credibility due to the general inaccuracy of this set of results.

When observing the theoretical data of U versus flow rate across all three exchangers, direct proportionality is shown. Unfortunately this trend cannot be verified by the results attained during this investigation, for a combination of systematic and random errors. Generally, this experiment suffers from a lack of information regarding the cleaning (or lack of it) in the exchangers as this is a major design consideration and a potential large layer of resistance. In addition, the experiment would be of more use if further data points were collated as it would help identify and verify correlations and trends within the results. In addition, there are equipment limitations that severely hamper the experimental process, such as the inability to maintain the required level of pressure for the steam supply and failure to increase the cold inlet to the shell & tube exchanger past 22 l/min. these issues should be resolved for future experimenters or a new experimental procedure should be established.

Conclusion
The aim of the experiment was to investigate the efficiency of three different types of shell & tube heat exchanger by examining the relevant temperatures across the system at steady state. Additionally, the relationship between the overall heat transfer coefficient U and the flow rate was evaluated. Ultimately the experiment managed to show represent the efficiencies of the shell & tube and spiral exchangers within a reasonable degree of accuracy, with respect to errors present in the system, as 56.8% and 39.1% respectively. The plate heat exchanger demonstrated a large discordance with the other two results by as the theoretical U value is significantly larger leading to a very small efficiency at 17.3%, however this result does not reflect the true value and it is recommended that a different theoretical approach should be taken. The experimental results offer no substantial basis to support the theoretical analysis with regards to the relationship between U and the flow rate.

References
1. Richardson, J.F.; Harker, J.H.; Backhurst, J.R. (2002). Chapter 3: Motion of Particles in a Fluid. In: Coulson and Richardson's Chemical Engineering Volume 2, (5th Edition). Elsevier. Equations 2 (p149), 3, 5, 6 (p155), 7 (p157), 8 (p156), 10, 11 (p157), 12 (p148), 13, 14, 15 (p167) Tables 3 (p157), 4 (p168) 2. S.B.A. Invent (2012). Fluid Mechanics: Bernoulli's Equation. Available at: http://sbainvent.com/fluid-mechanics/bernoulli's-equation.php#.UKqlOIfZbYQ last accessed: 19/11/2012 Figure 1

Appendices i. Nomenclature
This table shows the values used throughout this report and the units which apply.

ii. iii. iv. v. vi.

Example Example Example Example Example

You might also like