26Th Hexag Meeting

26TH HEXAG MEETING

QUEEN MARY UNIVERSITY LONDON, 23 OCTOBER 2007.

MEETING MINUTES

Introduction & Welcome

The meeting was hosted by Prof. John Rose of Mechanical Engineering at QMUL (). Approximately thirty members were present. John welcomed us with a brief history of the College, highlighting the recent plethora of name changes, and John also described the Faculty structure.

Technical Presentations[1]

The technical presentations were opened by Hans Zettler of HTRI, () who presented data on heat exchanger design margins. Hans firstly described HTRI, its history and structure, and the facilities for research, mainly based in the US in Texas. These included multi-purpose condensation and boiling units, the latter being used to look at flows in U bends. There was also a high temperature fouling unit, work on single phase laminar flow and crude oil fouling. Work on spiral heat exchangers was currently under way.

Moving on to the challenges for design margins, Hans mentioned TEMA fouling factors as an example, these having remained unchanged since around 1940. The philosophy emerging was firstly to add flexibility to practice in heat exchanger design and secondly to offer an alternative to fouling factors. This, for example, would necessitate the analysis of risks associated with adjusting design margins. The margin is an economic and business decision. Hans outlined the reasons for adding margins, and said that fluid properties were an important factor here. This was illustrated using real cases of the margin in terms of cross-sectional area added for fouling, and an extra 70-100% was typical.

There is uncertainty in the heat transfer coefficient (HTC). This could vary by 20-50% on the shell side of shell & tube heat exchangers. Hans said that if one was dealing with a non-fouling steam, fouling factors should not be used. If fouling was present, steams that foul most heavily should be pit on the tube side, where higher velocities can be used. The approach was illustrated using a gas conditioning train case study.

Weizhong Xiang of Hoare Lea Consulting Engineers () described his work on CFD modelling of the air flow and heat distribution within buildings. He highlighted Government diktats relating to carbon emissions from the built environment, including the fact that all its own property should be carbon neutral by 2012 and this will extend to new houses by 2016. Solutions include, for the UK, the use of biomass, geothermal heat pumps, wind/solar energy and natural ventilation. These were appropriate where ambients were typically in Summer 20-25oC and rarely exceeded 30oC.

Weizhong described modelling of natural ventilation in stacks and wind-driven ventilation. He said that the latter was the most challenging as flows were inconsistent and calculation was difficult. Critical components were a turbulence model, knowledge of wind profiles, the selection of the grid size and the extent of the calculation domain, which could be of the order of 1 km2 for a building. There are three ways to model turbulent flow commonly used, the RNG model being used in industry, as recommended by the EU COST programme. Wind profile calculations take into account roughness on the ground, e.g. features in a large city at/near ground level. Profiles are predicted from inlet regions to thousands of metres downstream, and examples of the profiles were shown.

The mesh scheme for 2D models was described, the domain extending to five times the building height. An example of a naturally-ventilated building was given – where the heat gain was 9.3 W/m3. The use of louvers in the walls aided ventilation. The flow/temperature patterns were shown for a building fully open inside and with floors. The actual conditions in Cardiff, with a 5 m/s wind, were superimposed on the model and the velocity and temperatures resulting were shown.

Nanofluids are an area of growing interest and Dongshen Wen of QMUL () reviewed the topic for us in the context of heat transfer intensification. He firstly defined them, highlighted their size-dependent properties and cites some applications, ranging from structures through biomedical uses to energy (combustion, oil recovery etc.). Dongshen said that most nano papers to date had concentrated on heat transfer intensification. He introduced the term ‘nanofluids engineering’ to describe a systematic view. This covered a range of topics that required addressing in any application, including multiscale modelling, synthesis, formulation, property measurement, application and environmental considerations.

The thermal conductivity comparison between water and a nanofluid was interesting: water was 0.6 W/mK while copper is 400 and carbon nanotubes is in the range 200 to 6000 W/mK. Thus a mixture of liquid plus nanoparticles has a greater conductivity than the liquid alone. Graphs of thermal conductivity as a function of particle volume fraction showed that in one particular case a 1% volume fraction of particles gave a 2.5 times conductivity enhancement. Also, dilute nanofluids have little effect on pumping power. Dongshen showed us a sample of a nanofluid that resembled skimmed milk. He said that formulation and characterisation of nanofluids were missing from most research, and remained a challenge. Viscosity and wetting features remain unknown and shear thinning of nanofluids has been observed. Some measured thermal conductivities have been much higher than those predicted. Several mechanisms have been proposed for this difference, including Brownian motion, nanoparticle clustering and the effect of particle shape. It was concluded that a fundamental understanding is needed of the heat and mass transfer between nanoparticles and the liquid in which they are located.

Dongshen showed data on forced convection, with alumina particles. Conductivity increased by 10% (cf water with 0.6-1.6% nanoparticles by volume) while the HTC increased by 40-50%. If carbon nanotubes were used the HTC could be up to 200% higher. In the case of boiling heat transfer with aluminium oxide and titanium oxide in water, one got 10-50% enhancement, but some such as Das in India have reported decreases. QMUL has measured a 20-40% increase. The interaction between the nanoparticles and the heat transfer surface is unclear.[2]

Joe Quarini of BristolUniversity () then updated us, in his inimitable style, on his work on ice pigging. Pigging, Joe said, was invented to aid hydrocarbon recovery, and he showed us various forms of ‘intelligent pigs’. Joe’s variant is based upon ice – his glacier analogy illustrated this neatly – and the flexibility of this form of pigging is that it can handle changes in pipe cross-section (to 24% in area) and 180o bends. Recent uses have been by Bristol Water for sand removal from pipes and, in the food sector, removing jam from pipes. With the aid of video clips Joe showed us these and other features of ice pigging, including the fact that an orifice plate does not inhibit the pipe filling up with ice.

The effect on pressure drop was shown with different ice fractions – at high volume flow rates the position seemed better – and the pig has been put through plate heat exchangers and a Spiraflo heat exchanger. The possibility of maldistribution in the plate unit was examined by using a see-through unit, again illustrated via a video.

The recent summer floods allowed a vivid demonstration of pigging of water pipes in Southwest England, a 75 mm diameter pipe of water with sand containment being fully cleared by injecting ice via a 25 mm secondary pipe into the mains. Joe also mentioned the use of ice pigs in the nuclear industry, where magnesium oxide slurry from the fuel elements could set in the pipes if not cleaned.

We returned to the small scale when Yuying Yan of NottinghamUniversity () introduced us to biomimetics and low adhesive and hydrophobic surfaces. Biomimetics is concerned with abstracting good design from nature, and examples used by Yuying to illustrate the phenomenon included the micro-roughness on the surface of a lotus leaf and the surface of a butterfly wing. The interesting feature of the latter was that the surface was normally hydrophobic, but if damaged it became hydrophilic.

Yuying listed several applications of biomimetics, but emphasised that reproduction was not just a matter of copying geometry – physical similarity was needed and surface tension had an important role to play in any model. He described the mathematical approaches to modelling and design and showed videos of bubble coalescence at the micro scale, including flow in microchannels. An interesting example was the ability of a ‘hydrophobic’ strip on a hydrophilic (wetting) surface could change the surface potential and in this case allow a drop to be separated into two parts, one each side of the strip. This could lead to self-cleaning surfaces and could be introduced in woven textile structures.

Another use of phenomena at these scales was electro-osmosis, which could be used to pump liquid through small channels/pores. This allowed Yuying to lead us into the applications in heat exchangers. Micro-electro-osmotic pumps, the use of surface electro-osmotic effects to reduce drag and the application of hydrophobic and/or hydrophilic surfaces for self cleaning, in plate heat exchangers, for fouling reduction and to reduce pressure drops.

David Reay () then updated us on a study carried out on behalf of the Carbon Trust into the potential for carbon savings through the use of process heat recovery. He began by outlining the aims of the study, as presented at the HEXAG meeting (25th) in Newcastle last year:

•Identify the overall savings potential for UK Industry from process heat recovery.

•Identify the savings potential for each significant industrial sector from process heat recovery.

•Identify process heat recovery technologies and estimate savings potential for each technology.

•For each technology, determine whether its use should be encouraged and supported by the Carbon Trust, either from the ECA scheme or possible other Carbon Trust support mechanisms that could be applied.

•

The final task included preparation of a strategy that the Carbon Trust might follow in order to implement PHR in industry in the most effective way.

• Analysed data suggest energy savings across 17 sectors of almost 900 ktoe/a (or 2.6% of energy use in identified sectors)
• ‘Top’ sectors for PHR are petroleum refining, food & drink, basic metals and chemicals, where the average percentage savings rises to 3.4%
• Based largely on ’conventional’ PHR technology – i.e. heat exchangers, but heat pumps have large potential

In the view of those consulted, including HEXAG members, during the study, the following points were made: ‘Best’ PHR units are expensive. Perceptions of fouling can hinder take-up. The question is not ‘can it be done?’ but ‘is it worth it?’ So some stimulation is needed. The UK resources for supporting large-scale take-up of some technologies are poor.

For the companies prepared to invest in making PHR units, the rewards, subject to sufficient backing from Government, could be substantial (as of course could be the carbon reductions). For example, there could be a market for over 4000 100 kW duty gas-gas heat exchangers and over 500 150 kW closed cycle heat pumps. Although direct stimulation of the heat pump market may be unlikely, except in the buildings sector where most emphasis seems to be these days, the heat pump, in David’s view, offered the greatest potential to improve process efficiency, unless the process is changed completely by, for example, process intensification.

However, the current lack of incentives for advanced heat exchanger research, as recommended by the Carbon Trust, was illustrated by quotations from the CT research landscape study published in 2005, in which advanced heat exchangers and process intensification were to be ‘reviewed periodically’[3]. Interestingly, £1 million was awarded by EPSRC to a research project on advanced heat exchangers, specifically two-phase heat transfer, in 2005 to a consortium represented at the HEXAG meeting.

Impromptu Presentations

The first impromptu was an update on work on internal coatings for heat exchangers by Chris Whelan of WDL Power (). Much of the work has been connected to automotive heat exchangers, in particular charge air coolers and exhaust gas recirculation units, which are being combined into a single unit – one such system is now used on DAF trucks. A problem, Chris said, was soot deposition (particle size less than 5 microns) inside these exchangers, and sol-gel coatings might reduce soot adherence as well as protecting against corrosion. Research has examined how soot detaches from surfaces, and the effect of soot build-up on exchanger performance has been modelled. The concept has been patented and demonstrated on a car.

Chris said that they now have a solution to putting soot-laden gases through an aluminium heat exchanger, and internal coatings have potential. Chris said that R&D is now needed on coating science and related areas.

Bart Hallmark of CambridgeUniversity() re-introduced us to polymer microcapillary films (MCFs) which could be used for heat transfer and also in intensified unit operations, as he first revealed at the Newcastle HEXAG meeting. The polymer film ‘ribbon’ contains holes in the form of longitudinal capillaries of down to 7 microns in diameter.

In the new concept, Mark showed a multi-layer countercurrent polymer film heat exchanger which had 9 films on the cold side, 10 on the hot side and handled 380 ml/min of fluid. With a duty range of 350-700 W, (when the flow might reach 900 ml/min), the advantages included very low residence times allowing temperature changes of 80 K/sec and one could envisage 100-150 K/sec changes. Heat transfer coefficients were of the order of 600 W/m2K.

The final talk was by Jacqueline Barber, a PhD student at EdinburghUniversity (). The thrust of her research is into two-phase boiling and flow instabilities in a microchannel. The outcome of her research could find application in the design of ink jet printers, micro-electro-mechanical devices (MEMS) and compact heat exchangers. Jacqueline described previous work in this area and gave us details of her test facility that included a 4000 frames/sec camera and a syringe pump for fluid injection. An infra-red camera was used to examine flux/temperature changes and this could operate art 50 frames/sec. The microchannel that was used for the measurements had a hydraulic diameter of 727 microns, was rectangular in cross-section and used n-pentane as the working fluid.

A comprehensive tour of the laboratories, concentrating on condensation research, was laid on, and thanks are due to John Rose, Adrian Briggs, Huasheng Wang, Dongshen Wen and their postgraduates for their hospitality.

***************

Minutes prepared by David Reay, 26 October 2007.

1

[1] Authors are invited to send me their presentations for inclusion on the HEXAG web site.

[2] See also proceedings of the recent UK Heat Transfer Conference for data on nanofluids and boiling heat transfer.

[3] CTC511, The Carbon Trust, 2005. See the CT web site to download the document.