Fuel Processing Technology 119 (2014) 32–40

Contents lists available at ScienceDirect

Fuel Processing Technology

j ourna l homepage: www.e lsev ie r .com/ locate / fuproc

Development of a novel solids feed system for high pressure gasification

J.M. Craven a,⁎, J. Swithenbank a, V.N. Sharifi a, D. Peralta-Solorio b, G. Kelsall c, P. Sage c

a Energy and Environmental Engineering Research Group (EEERG), Department of Chemical and Biological Engineering, University of Sheffield, Mappin Street, Sheffield S1 3JD, UKb E.ON New Build and Technology Limited, Technology Centre, Ratcliffe-on-Soar, Nottingham NG11 0EE, UKc BF2RA, Gardner-Brown Limited, Calderwood House, 7 Montpellier Parade, Cheltenham GL50 1UA, UK

⁎ Corresponding author at: Department of ChemicaUniversity of Sheffield, Mappin Street, Sheffield S1 3JD,fax: +44 114 222 7501.

E-mail address: james.craven@sheffield.ac.uk (J.M. Cra

0378-3820/$ – see front matter © 2013 Elsevier B.V. All rihttp://dx.doi.org/10.1016/j.fuproc.2013.10.016

a b s t r a c t

a r t i c l e i n f o

Article history:Received 3 July 2013Received in revised form 28 August 2013Accepted 24 October 2013Available online 16 November 2013

Keywords:Feeding systemGasificationSolid fuelsHigh pressure processes

The Hydraulic Lock Hopper (HLH) embodies a high pressure dry feed system that uses water as an incompress-ible fluid to bring about compression. No pressurising gas is required, so commonly used inert gases such asnitrogen and carbon dioxide are conserved. The HLH has successfully demonstrated the feeding of solid fuelssuch as wood pellets to pressures as high as 25 barg in two modes of operation. Energy requirements of15.51 kJ/kg (Mode 1) and 20.61 kJ/kg (Mode 2) have been recordedwhich translate to significant energy savingsof 81.9% and 75.9% compared to conventional lock hoppers. Energy savings have been projected to increase forMode 2where lock gas contaminationwith syngas takes place, and themassflow rate has been shown to operateindependently of pressure varying between 2 and 2.5 tonnes/day. The HLH has also been shown to have anegligible effect on the fuel moisture content with moisture content increases being recorded to be consistentlyless than 1 wt.%.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

In recent years there has been an increased drive to produce energyfrom renewable sources. However, conventional fuels such as coal, oiland gas still dominate the landscape in terms of energy productionand account for approximately 87% of the world's total energy supply[1]. Gasification poses as a technology that can provide both a carbonfree power supply where biomass is used, and a more efficient powersupply using conventional fuels when integrated as a combined cycle(IGCC). However, in order to operate combined cycle processeseffectively, the feed stream has to be pressurised prior to use in thegas turbine stage. This presents two options for operation: either thegasifier operates at atmospheric pressure and the syngas produced ispressurised to the desired pressure for operation post gasification, orthe gasifier operates at elevated pressure to produce a syngas productat pressure ready to be utilised in a gas turbine. The latter route bothpresents savings in the energy required for syngas compression andallows a reduction in equipment size [2–4]. It is for this reason thatpractically all modern gasifiers operate at pressures above that of theatmosphere, typically between 25 and 40 bar [4,5].

However, at the expense of increased operating efficiency comesincreased process complexity; not least with the way in which thesolid feedstock is introduced to the process at pressure. Where gasifica-tion at elevated pressure is concerned, the feeding system has beencited as one of the most common causes of process downtime [6,7].

l and Biological Engineering,UK. Tel.: +44 1142 224 910;

ven).

ghts reserved.

Added to this the fact that current feeding systems are typically ineffi-cient and present a burden on the overall efficiency of the gasificationprocess highlights the need for alternative devices to be developed.

1.1. Scientific background

High pressure solids feeders can be split into six categories: rotaryvalves, lock hoppers, plug-forming feeders, piston feeders, dynamicfeeders and slurry feeders [8–15]. Amongst these, lock hoppers andslurry feeders are the most common and established feed systems;however, neither system is without its operational drawbacks [3,4].

The principal drawback of lock hoppers is their inefficient operationdue to a large reliance on gas used for pressurisation. The largest portionof the power required to facilitate the feeding operation is used in thisstep and is ultimately wasted during the depressurisation stage[12,16]. Both the power used for pressurisation and the gas used arewasted due to pressurisation cycling, and losses are seen to increasedramatically with increasing pressure.

Slurry feeders look to counter issues regarding inefficient feedingthrough the adaptation of conventional liquid pumps. However, thelarge amount of water used to make up the slurry, typically 60–70 wt.%solids in the case of coal [9] and 10–15 wt.% solids in the case ofbiomass [17], poses as the main drawback in their operation. Gasifi-cation only requires a small portion of the water present in slurriesfor steam formation and use in the water-gas shift reaction, and theremaining water or moisture content serves to decrease the overallefficiency of the gasification process. A large amount of energy isrequired to vaporise any excess moisture present in the fuel andthis in turn increases oxygen consumption and leads to decreasedcold gas efficiencies. In practice, dry fed gasifiers have a 20–25%

33J.M. Craven et al. / Fuel Processing Technology 119 (2014) 32–40

lower oxygen consumption for this reason alone [4]. The disadvantage ofslurry feeding is particularly felt where the gasification of biomass andlow rank coals is concerned, as such fuels inherently contain high levelsof moisture. It is generally accepted that combustion and gasificationreactions cannot proceed where moisture contents exceed 60–65 wt.%[18,19]. Given that the internal moisture content of a fuel does notcontribute towards the moisture required to make up a slurry leads tothe conclusion that feeding biomass and low rank coals as a slurry isnot a viable option [4].

A number of feed systems within the remaining categories havebeen developed to counter the drawbacks stated and are present inthe literature [8–10,15]. However, the majority of devices developedhave failed to reach a commercial stage, leaving lock hoppers and slurryfeeders as the twomost common types of feed system in the gasificationindustry. The objective of this study is to develop a solids feed system forefficiently feeding solid fuels to high pressure processes, and to counterthe common drawbacks experienced in current systems.

2. Materials and methods

2.1. Hydraulic Lock Hopper

The Hydraulic Lock Hopper (HLH) takes the form of a new lockhopper that utilises water as an incompressible fluid. Water is used tominimise the amount of work required in the compression stage ofthe feeding operation by displacing gas at high pressure. Therefore,the bulk of the compression work is carried out by the water and inturn the pump used to introduce the water into the high pressureenvironment. A high pressure water pump provides the only mechani-cal action during the feeding operation and is used in place of a com-pressor as found in conventional gas compression cycling. A schematicdiagram of the HLH and the experimental rig is shown in Fig. 1.

The HLH consists of two hoppers equal in volume, a high pressurecollection vessel and a high pressurewater pump. As the feeding systemoperates as a batch process, the role of the collection vessel is to act as adestination to be fed to at pressure. In reality the collection vessel canact as a “live bottom” feeder using a variety of established feed systemsto provide continuous feeding of solid fuel to the desired downstreamprocess. Feeding devices such as metering bins and screw feedersare routinely used in such scenarios and have been demonstratedspecifically by the lock hopper developed by Thomas R.Miles ConsultingEngineers for such a purpose [13,14].

The atmosphere and the top hopper are separated by a ball valve(V1), the top and bottom hopper are separated by a ball valve (V2)and similarly the bottom hopper and the collection vessel are separatedby another ball valve (V3). Further to this, the top and bottom hopperare connected by an external pipe outside of the main body of thefeeding system which contains an isolation valve (V4). This connectingpipe is used for pressure equalisation and recompression. Anotherconnecting pipe is fitted to the top hopper which contains three valves(V5, V6 and V7).Water is fed from the high pressure water pump to thetop hopper via V5 and V6 and is drained via V5 and V7 to awater collec-tion chamber. A third and final pipe containing a valve (V8) is added tothe top hopper in the form of a syphon which serves to reduce thevolume of water retained in the top hopper to a minimum after thewater drainage stage.

2.2. Modes of operation

TheHLH can be operated in two separatemodes:Mode 1 andMode 2.Mode 1 requires a volume of water to be pumped to the top hopperthat is approximately equal to that of the void space present betweenthe fuel; this typically varies between 40 and 60% for most solid fuels[20]. Assuming a voidage of 50% would require the top hopper to behalf filled with water to compress the fuel to the desired operatingpressure. In Mode 1, the HLH begins its cycle with the top hopper

at atmospheric pressure, the bottom hopper at the desired operatingpressure, and both hoppers empty of fuel. All valves start in the closedposition, and therefore the opening of V2 brings about pressureequalisation between the top and bottom hopper prior to the pumpingof water in the compression stage of the feeding operation.

Mode 2 is more energy intensive than Mode 1 and requires the tophopper to be completely filled with water; this minimises the volumeof gas at pressure vented to the atmosphere during the decompressionstage. Contrary to Mode 1, V3 is left in the open position at all timesduring Mode 2 as to avoid a considerable over pressure in the bottomhopper; in Mode 2 the bottom hopper only acts to increase the volumeof the collection vessel. A slight net increase in the overall processpressure is encountered during this mode of operation due to thevolume taken up by the fuel fed. However, such an increase is consid-ered to be negligible when the overall volume of the downstreamprocess being fed to is taken into consideration. Like Mode 1, the HLHbegins its cycle with the top hopper at atmospheric pressure, bothhoppers empty of fuel, and all valves aside from V3 in the closedposition.

The reason for the hoppers of equal volume (top, H1 and bottom,H2) becomes apparent when the compression work required by theHLH is examined alongside a conventional single lock hopper with avarying volume ratio of H2:H1 (Fig. 2). In the case ofMode 1, the energysaving increases with a decreasing volume of the vessel being fed to,which brings about a minimum volume equal to the hopper locatedimmediately upstream. The opposite of this is found to favour Mode 2.Therefore, leaving V3 open provides a more efficient way in which tooperate Mode 2 as the bottom hopper acts to increase the total volumeof the collection vessel.

3. Results and discussion

3.1. Cycle analysis and operation

The principle of the HLH was assessed using 6 mm wood pelletsmade available by CPL Distribution Ltd. The pellets used were producedfrom chemically untreated residues from the wood processing industryand to satisfy the standard ENplus A1.

The HLH is a batch feeding system and therefore the geometry of thesystem governs the mass flow rate and how it is operated. A maximumof three batches of fuel can be fed across the pressure boundary duringone complete cycle (defined as system pressurisation, fuel feedingand system depressurisation) and therefore results relating to cycleoperation are presented as such.

Fig. 3 provides an overviewof the internal pressure and temperaturereadings during operation in a 25 barg case forMode 1 andMode 2. Thetemperature changes observed in the top hopper, bottom hopper andcollection vessel are due to the changes in pressure brought aboutthrough the compression and expansion of air. The expansion andcompression in this instance represents a polytropic process approxi-mating an isothermal process as temperature changes are observedwith little net change in the overall temperature over extended periodsof time.

In the case of Mode 1, temperature variations are recorded by TC3and TC2 in the top and bottom hopper respectively, while the tempera-ture in the collection vessel is seen to stay constant. The temperaturechanges observed in this case can be attributed to the large changes inpressure observed in the top and bottom hopper during operation. Sim-ilarly, the constant temperature reading in the collection vessel denotesa stable and constant pressure as is observed.

In the case ofMode 2, temperature variations similar those observedinMode 1 are found. However, temperature variations are also observedin the collection vessel and can be broadly compared to those found inthe bottom hopper. During Mode 2, V3 is left open throughout opera-tion and the bottomhopper acts to increase the volume of the collectionvessel. Larger temperature variations are recorded by TC2 in the bottom

Fig. 1. A schematic of the HLH and the constructed experimental rig.

34 J.M. Craven et al. / Fuel Processing Technology 119 (2014) 32–40

hopper compared to TC1 in the collection vessel due to the relativeproximity of TC2 to the top hopper. This lag in temperature variationis observed between the bottom hopper and collection vessel due tothermal losses and the distance between each respective thermocouple.Further to this, the maximum temperature recorded by TC3 in the tophopper during operation is seen to be considerably higher in Mode 2than in Mode 1. This is accounted for by the larger volume of the vessel

located immediately downstream of the top hopper in Mode 2, and inturn the larger change in pressure generated.

3.2. Mass flow rate and cycle time

The HLH was operated using 4 kg batches of 6 mm wood pellets.The fuel voidage was determined to be approximately 46.5% which

Fig. 2. Effect of H2:H1 ratio on energy saving compared to a conventional single lockhopper (fuel voidage = 50%, pressure = 25 barg).

Fig. 4.Mass flow rate operating in both Mode 1 and Mode 2.

35J.M. Craven et al. / Fuel Processing Technology 119 (2014) 32–40

translates to a void space present in the top hopper of approximately61%. The difference between the two values for void space is accountedfor by the head space present above the fuel level in the top hopper priorto feeding. Such a head space is required to ensure reliable and stableflow of the material from vessel to vessel and to avoid operationalproblems such as solids holdup and valve failure.

A general trendof decreasingmassflowratewith increasing operatingpressure is observed from Fig. 4 further to Mode 1 N Mode 2 at all

Fig. 3. Temperature and pressure variations during

operating pressures. However, when the overall operating cycle perbatch of fuel fed is examined, the three fundamental stages offeeding (gravity feeding, water compression and water drainage)are seen to be broadly constant with pressure. Contrary to what isexpected, the time taken to drain water from the top hopper afterfeeding is not seen to vary dramatically with pressure. Similarly,the time taken to feed water via the high pressure water pump isnot seen to dramatically change with pressure in Mode 1, withMode 2 highlighting a marginal increase with increasing pressure.

operation in Mode 1 and Mode 2 at 25 barg.

Table 1Overview of the feeding procedure in Mode 1 and Mode 2.

Pressure (barg) 10 15 20 25

Mode 1 Gravity feeding (s) 18 ± 2 18 ± 2 18 ± 2 18 ± 2Water compression (s) 55 ± 2 55 ± 2 54 ± 2 56 ± 2Water drainage (s) 29 ± 1 30 ± 3 30 ± 1 30 ± 1Valve activity (s) 36 ± 7 42 ± 10 49 ± 15 50 ± 9Total (s) 138 ± 8 145 ± 11 151 ± 15 154 ± 9

Mode 2 Gravity feeding (s) 12 ± 1 12 ± 1 12 ± 1 12 ± 1Water compression (s) 69 ± 1 71 ± 1 72 ± 1 74 ± 1Water drainage (s) 33 ± 2 32 ± 2 34 ± 2 33 ± 1Valve activity (s) 30 ± 5 30 ± 9 34 ± 3 35 ± 4Total (s) 144 ± 6 145 ± 9 152 ± 4 154 ± 4

36 J.M. Craven et al. / Fuel Processing Technology 119 (2014) 32–40

As seen from Table 1, it is the variability of manual valve operationthat accounts for the general decreasing trend of mass flow ratewith increasing operating pressure. It can therefore be stated thatmass flow rate is not directly affected by operating pressure wherean automated process is concerned and is independent of pressure.

Fig. 5. Energy use of the HLH and a conventional single and dual lock hopper (A). Theoreticaland experimental energy saving of the HLH compared to a conventional lock hopper (B).

3.3. Energy consumption

3.3.1. Mode 1 and Mode 2To assess the energy requirement of the HLH the electrical power

requirement of the high pressure water pump must be reviewed. Aclear trend of increasing power drawn with operating pressure can beseen for both modes of operation and is shown in Table 2. Whencombinedwith the results displayed for cycle time in Table 1 and specif-ically the time taken for compression, a clear increase in the energyrequirement to feed a batch of fuel is observed.

Fig. 5(A) shows a general trend of increasing energy requirement tofeed a batch of fuel with increasing back pressure with Mode 1 b Mode2 at all operating pressures.

Mode 1 consistently has a lower energy requirement per unit massdue to the fundamental difference between the twomodes of operation.Mode 1 requires a volume of water to be pumped that is approximatelyequal to the void space present within the top hopper containing fuelprior to feeding, and Mode 2 requires a volume of water equal to thevolume of the top hopper.

The volume ofwater pumped to the top hopper during the compres-sion stage in both modes of operation was recorded to be consistent atall back pressures. An average volume of 5560 ml was recorded to beused during operation in Mode 1 which approximately corresponds tothe total voidage present in the top hopper (65%). However, in thecase of Mode 2 the volume of water pumped was not recorded to beequal to the total volume of the top hopper as is required to minimiseand avoid the waste of gas at high pressure. The volume of the top hop-per when taken as a single entity is approximately equal to 7590 ml.However, the volume of the top hopper inclusive of pipe work andfittings occupied by gas at high pressure is approximately equal to8520 ml. The volume of water pumped when operating in Mode 2 ispredetermined and set at 7500 ml in order to avoid the carryover of

Table 2Power drawn during the compression stage for operation in Mode 1 andMode 2 and as aconventional single and dual lock hopper.

Pressure (barg) Power drawn (W)

Mode 1HLH

Mode 2HLH

Conventional lockhopper

Dual lockhopper

10 927 ± 31 974 ± 19 1514 ± 35 1554 ± 3015 998 ± 57 1012 ± 21 1524 ± 40 1521 ± 3920 1091 ± 33 1061 ± 21 1600 ± 38 1534 ± 3925 1104 ± 24 1123 ± 19 1590 ± 32 1528 ± 32

water into the bottom hopper/collection vessel. This minimises thecontamination of the fuel being fed with water, and in the case ofmoisture unstable fuels such as of pelletised biomass, maintains struc-tural stability. As a consequence, phenomena such solids holdup andvalve failure are minimised.

3.3.2. Comparison to current technologiesFurther to the energy use of the HLH being assessed for operation in

Mode 1 andMode 2, the feeding systemwas operated as a conventionalsingle and dual lock hopper in order to gauge a direct scale comparisonto conventional units. The feeding system was operated under thesame conditions as used for Mode 1 and Mode 2, with 4 kg batches offuel being fed. In place of the high pressure water pump, the threestage compressor used to pressurise the system as a whole was usedto simulate pressure cycling found in conventional single and duallock hoppers.

Fig. 5(A) highlights a far greater trend of increasing energy use withincreasing pressure for both a conventional single and dual lock hoppercompared to Mode 1 or Mode 2. Although a dual lock hopper is shownto require approximately 45–50% (10–25 barg) the energy requiredby a conventional lock hopper, energy savings of approximately 81.9%and 75.9% are recorded for the HLH operating at 25 barg in Mode 1and Mode 2 respectively. At lower pressures (≤10 barg) a dual lockhopper is seen to become more competitive with the HLH with anaverage energy use of 17.03 kJ/kg at 10 barg. This compares favourably

37J.M. Craven et al. / Fuel Processing Technology 119 (2014) 32–40

to Mode 1 (12.73 kJ/kg) and Mode 2 (16.90 kJ/kg). However, aspressure is increased, the energy use of the HLH is seen to be far lowerthan either a dual or a conventional lock hopper. Energy savings of ap-proximately 63.7% and 51.7% are recorded for Mode 1 and Mode 2operating at 25 barg respectively over a dual lock hopper. Such energysavings compared to a conventional single and dual lock hopper areprojected to increase for higher operating pressures.

Analysing the data recorded for power it can be seen from Table 2that unlike Mode 1 and Mode 2, there is no clear relationship betweenpower drawn by the compressor and operating pressure for either aconventional or a dual lock hopper. Power drawn is seen to be approx-imately constant regardless of pressure and ranges between 1500 Wand 1600 W with an average power drawn of 1550 W. The trend ofincreasing energy use with increasing pressure recorded thereforereflects a variable compression time and in turn a general trend ofdecreasing mass flow rate with increasing operating pressure. This iscontrary to a variable power drawn and approximately constant massflow rate when operating in Mode 1 and Mode 2 of the HLH.

Further to experimental work, the compression work required byboth a conventional lock hopper and the HLH operating in Mode 1and Mode 2 was calculated using Eqs. (1) and (2) for an isothermalcompression and adiabatic compression respectively:

W ¼ P1V1 � lnP2

P1

� �ð1Þ

W ¼ P1V1 �γ

γ−1� P2

P1

� �γ−1=γ−1

!ð2Þ

Fig. 6. Energy use taking into account varying deg

where W is the work done (J), P1 is the initial pressure (Pa), P2 is thefinal pressure (Pa) and γ is the ratio of specific heats (Cp/Cv) equal to1.4 for air [21]. Eqs. (1) and (2) provide an upper and lower limit forthe energy saving generated by the HLH. Fig. 5(B) highlights howthe experimental results relating to energy saving compare to thosedetermined theoretically. It can be seen from Fig. 5(B) that experimen-tal work fitswell with theory, with bothMode 1 andMode 2 lyingwith-in the upper and lower limits calculated given the stated error.

3.3.3. Effect of lock gas contamination on energy useAs one of the immediate benefits of the HLH is that the volume of

lock gas vented to the atmosphere is reduced (Mode 1) and largelyavoided (Mode 2), it is timely to analyse the energy use taking intoaccountwaste. By assuming varying volumepercentage contaminationsof lock gas with syngas and assuming that the lock gas is vented andnot recovered allows inherent energy wastage to be factored into theenergy requirement of the feeding system.

Although Mode 2 of the HLH ultimately seeks to negate lock gasventing, in reality there will still be a small volume that cannot bedisplaced with water during the compression stage. In this case theaverage volume displaced bywater for Mode 1 andMode 2was record-ed to be 5560 ml and 7530 ml respectively. With a total volume of thetop hopper being equal to 8520 ml, the volume percentage of lock gasat pressure vented is equal to 34.7% and 11.6% for Mode 1 and Mode 2respectively. However, lock gas contamination affects a conventionaland dual lock hopper far more dramatically, with a conventional lockhopper venting a full volume of gas at pressure after feeding. Fig. 6displays a combination of the experimental data generated for Mode 1,Mode 2, a conventional lock hopper and a dual lock hopper with

rees of lock gas contamination with syngas.

38 J.M. Craven et al. / Fuel Processing Technology 119 (2014) 32–40

a theoretical energy waste due to venting. A medium calorific valuesyngas of 15 MJ/Nm3 is assumed with a temperature of 1000 °C [5].The lock gas contaminatedwith syngas is assumed to expand to standardtemperature and pressure (STP) and the volume at STP is calculated via amolar basis.

It can be seen in Fig. 6 that lock gas contamination serves to increasethe total energy use of all cases and thatMode2 is the least affected case.At an operating pressure of 25 barg andwhere a lock gas contaminationof 12.2 vol.% takes place the overall energy use of Mode 1 is approxi-mately equal to that of Mode 2. Therefore, it can be said that wherelock gas contamination is N12.2 vol.%, Mode 2 is the most efficientroute for feeding.

Ultimately, it can be said thatMode 2 is themost favourablemode ofoperation where high lock gas contamination with syngas takes place,withMode 2 N Mode 1 N dual lock hopper N conventional lock hopper.This trend is projected to increase further where syngas streams withhigher calorific values are produced.

3.4. Effect of the HLH on fuel moisture content

Moisture content was assessed assuming that any mass increaseacross the pressure boundary was due to the uptake of water by thefuel. The mass of the fuel was recorded before and after feeding todetermine any moisture content increase generated by the HLH. Theprimary system pressure boundaries were imposed by 3 inch bore ballvalves (V1, V2 and V3) and two valve material types were tested:chrome plated brass and stainless steel (AISI 316) with materialhardness' of 80–95 HB and 150–190 HB respectively.

The HLH was run under the same conditions using both valvematerials with the order of testing being Mode 1 followed by Mode 2.Due to the higher pressure rating of the stainless steel ball valves(PN69) compared to the chrome plated brass valves (PN25), the HLHwas operated between 10 and 25 barg using the stainless steel valvesand 5–20 barg using the chrome plated brass valves, both with 5 barincrements.

It can be seen from Table 3 that smaller moisture content increasesfor operation in Mode 1 are observed where the chrome plated brassvalves were used. Comparatively, it is seen that for operation in Mode2, smaller moisture content increases are observed where the stainlesssteel valves were used. However, when the operation of the HLH as awhole is examined and the order in which the modes were run isreviewed (Mode 1 followed by Mode 2) it can be seen that the harderstainless steel valves provide a more consistent and reliable operation.Results for moisture content increase are seen to be independent ofoperating pressure in both modes of operation and are consistentlyb1 wt.%. Results therefore indicate that such increases are broughtabout due to residual moisture being present in the top hopper afterfeeding a previous batch of fuel rather than water passing through thevalve. Results where the chrome plated brass valves were used indicatethat this is also likely in the case of Mode 1. However, the valves areseen to degrade over time and give rise to water passing through the

Table 3Fuel moisture content increase with different valve materials.

Pressure (barg) Moisture content increase (%)

Chrome plated brass Stainless steel

Mode 1 Mode 2 Mode 1 Mode 2

5 0.06 ± 0.04 2.7 ± 0.2 – –

10 0.04 ± 0.02 1.74 ± 0.77 0.37 ± 0.25 0.54 ± 0.2115 0.10 ± 0.08 3.39 ± 0.47 0.27 ± 0.03 0.81 ± 0.2220 0.09 ± 0.07 4.00 ± 1.19 0.31 ± 0.12 0.22 ± 0.0125 – – 0.30 ± 0.07 0.50 ± 0.37

valve; as a consequence, higher increases in fuel moisture content aregenerated.

Examining both sets of valves after operation shows increasedmaterial wear and degradation in the case of the chrome plated brassvalves compared to the stainless steel valves. It can therefore be statedthat in order to cope with the harsh dusty environments imposedby the feeding of solids, harder material types are more suitable. Theselection of ball valves as the valve type in this instance was largelydown to scale and manual ease of use. In reality and in larger settings,knife gate valves are most suitable and have been demonstrated inconventional lock hopper designs to counter operational issues suchas solids holdup and valve failure [11,22].

Further to observing the physical wear of the valves, the level ofvalve degradation was assessed through air leakage measurements.

Fig. 7 shows that air leakage was found to take place through bothsets of valves and leakage rates to be an order(s) of magnitude greaterwhere the chrome plated brass valves were used. A general trend ofincreasing air leakage rate with increasing pressure was recorded inall cases, with V3 being shown to be the worst affected in the case ofthe chrome plated brass valves. Comparatively, V2 was shown to bethe worst affected in terms of air leakage where the stainless steelvalves were used, but by a far smaller margin. This is with the exceptionof the 25 barg case where air leakage is recorded to be dramaticallylarger through V2 than at previous operating pressures. Such air leakageresults are indicative of the results relating to fuel moisture contentincrease and serve to compound the assumption that increases inmoisture content where the chrome plated brass valves were used is

Fig. 7.Air leakage throughV2andV3using chromeplated brass and stainless steel as valvematerials.

39J.M. Craven et al. / Fuel Processing Technology 119 (2014) 32–40

due to water passing through the valve. Similarly, this also agrees wellwith the results recorded using the stainless steel valves, as such lowair leakage rates indicate little/no water to be passing through thevalves. It can therefore be said that increases in fuel moisture contentduring operation with the stainless steel valves are brought aboutthrough the absorption of residual moisture contained in the tophopper.

3.5. Industrial application

Awide range of industries can benefit from using the HLH in place ofconventional systems, in particular the paper and pulping industry,power generation industry and chemical industry [4,11,15,23,24]. Oneof the key advantages of the HLH over competing systems is its abilityto be retrofitted to existing lock hoppers at a relatively low cost andwithout the need for significant changes to be made to the overalldesign. The only materials and equipment required for retrofit arehigh pressure pipe work/fittings and a high pressure water pump.Water cleanup is required if water is to be re-used by the system as tominimise damage to the pump. However, water cleanup does not pres-ent a significant issue and can be brought about through standardwatertreatment methods (filters) at a low cost.

As the HLH can be implemented through retrofitting existing lockhoppers, one of the key advantages conventional lock hoppers haveover competing feed systems is enabled — feedstock flexibility. Theparticle size of the material being fed is an important factor in thefeeding process and affects the reliability of the majority of feedsystems. As the HLH is comprised of a series of hoppers operatingin series, well established design procedures used to design conven-tional lock hoppers are able to be employed; chiefly is the designprocedure developed by Jenike [25]. Therefore, the HLH is able tohandle a wide range of materials through custom design.

As previously stated in Section 3.3.2, the HLH highlights a clearadvantage over conventional lock hopper systems, generatingenergy savings of up to 81.9% for Mode 1 and 75.9% for Mode 2when operating at 25 barg. However, it is important to view this inthe context of an overall process. Energetically, it is convenient toassess the energy requirements of such feed systems alongside theenergy content of the fuel being fed. The energy requirements ofMode 1, Mode 2 and a conventional lock hopper translate to approx-imately 0.082%, 0.11% and 0.45% the energy content respectivelywhere a higher heating value (HHV) of 18.90 MJ/kg is assumed foran average biomass fuel [26]. Similarly, the energy requirementstranslate to approximately 0.048%, 0.064% and 0.27% the energy con-tent where a HHV of 32 MJ/kg is assumed for an average bituminouscoal [4]. This is of particular interest where large scale, large through-put processes are concerned.

A 4 GW coal gasification plant with an overall efficiency of 48% [4]would require around 22,500 tonnes/day of coal to be processed. Toattain the same power rating with biomass, approximately 1.7 timesthe amount of coal would be required to be converted. This translatesto approximately 38,250 tonnes/day of biomass. Table 4 provides anoverview of the potential fuel, energy, carbon emission and cost savings

Table 4Fuel, energy, carbon emission and economic savings generated by the HLH over a conven-tional lock hopper for a 4 GW gasification plant.

Saving (per day) Coal Biomass

Mode 1 Mode 2 Mode 1 Mode 2

Fuel (tonnes) 102.5 95.0 295.0 273.5Energy (MWh) 437.2 405.4 743.3 689.1CO2 (tonnes) 293.1 271.7 519.1 481.3Cost ($) 27,983 25,943 47,571 44,104

the HLH would generate if used in place of a conventional lock hopper.Cost estimates are assumed to be generated through electricity savingsusing $64/MWh (£40/MWh) as an average price per unit of electricityproduced on site [27], and CO2 savings assume a fuel fixed carboncontent of 48 wt.% for biomass [26] and 78 wt.% for coal [4].

4. Conclusions

The Hydraulic Lock Hopper has demonstrated the feeding of 6 mmwood pellets to pressures as high 25 barg in two key modes of opera-tion. Energy requirements of 15.51 kJ/kg and 20.61 kJ/kg have beenrecorded for operation in Mode 1 and Mode 2 respectively operatingat 25 barg, and energy savings of 81.9% (Mode 1) and 75.9% (Mode 2)compared to conventional lock hopper systems have been achieved.The mass flow rate through the HLH has been shown to functionindependently of operating pressure, ranging between 2 and 2.5 -

tonnes/day for an automated process, and the HLH has been shown tohave a negligible effect on the fuel moisture content with moisturecontent increases recorded being consistently b1 wt.%.

Acknowledgements

The authors gratefully acknowledge the Engineering and PhysicalScience Research Council (EPSRC) and the Biomass and Fossil FuelsResearch Alliance (BF2RA) for their financial support. Special thanksgo to Greg Kelsall (BF2RA), Peter Sage (BF2RA) and Dr David Peralta-Solorio (E.ON) for their technical support and to Mike O'Meara andDavid Palmer (University of Sheffield) for their assistance in theconstruction of the experimental rig.

References

[1] BP, BP Statistical Review of World Energy June 2013, BP, London, 2013.[2] H. Boerrigter, R. Rauch, Review of Applications of Gases from Biomass Gasification,

ECN Biomass, Coal and Environmental Research, Holland, 2006.[3] J. Phillips, Different Types of Gasifiers and Their Integration with Gas Turbines,

US-DOE, NETL, Electric Power Research Institute, Advanced Coal Generation,2006.

[4] C. Higman, M. Van Der Burgt, Gasification, Gulf Professional Publishing, 2008.[5] J. Rezaiyan, N.P. Cheremisinoff, Gasification Technologies: A Primer for Engineers

and Scientists, CRC, 2005.[6] D.L. Bonk, A. Hay, Stamet Solids Pump Feeds Coal into 210 psig in a DOE Supported

Project, Power-Gen America, Anaheim, California, 1995.[7] S.J. Clayton, G.J. Stiegel, J.G. Wimer, Gasification Technologies: Gasification Markets

and Technologies — Present and Future: An Industry Perspective, US-DOE, NETL,2002.

[8] Jet Propulsion Laboratory, Proceedings of the Conference on Coal Feeding Systems,U.S. Energy Research and Development Administration, California Institute ofTechnology, Pasadena, California, 1977.

[9] J.R. Hawrych, High pressure coal feeding systems for coal conversion plants,International Gas Research Conference, Worley Engineering Limited, 1981.

[10] B.H. Levelton, Status of Biomass Feeder Technology, Canadian Forestry Service, 1982.[11] A.R. Guzdar, A.C. Harvey, Feeding coal and biomass into high pressure reactors,

Biomass to Methanol: A Specialists' Workshop, US-DOE, Durango, Colorado,1982.

[12] A. Rautalin, C. Wilén, Feeding Biomass into Pressure and Related Safety Engineering,Technical Research Centre of Finland (VTT), 1992.

[13] C. Wilén, A. Rautalin, Handling and feeding of biomass to pressurized reactors:safety engineering, Bioresource Technology 46 (1993) 77–85.

[14] K.R. Cummer, R.C. Brown, Ancillary equipment for biomass gasification, Biomass andBioenergy 23 (2002) 113–128.

[15] M.L. Swanson, M.A. Musich, D.D. Schmidt, J.K. Schultz, Feed System Innovation forGasification of Locally Economical Alternative Fuels (FIGLEAF), DE-FC26-00NT40904US-DOE, NETL, Pittsburgh, PA, 2003.

[16] A. Van Der Drift, H. Boerrigter, B. Coda, M.K. Cieplik, K. Hemmes, R. Van Ree, H.J.Veringa, Entrained Flow Gasification of Biomass, ECN Biomass, Holland, 2004.

[17] E.J. Berglin, C.W. Enderlin, A.J. Schmidt, Review and Assessment of Commer-cial Vendors/Options for Feeding and Pumping Biomass Slurries for Hydro-thermal Liquefaction, US-DOE, Pacific Northwest National Laboratory, 2012.(DE-AC05-76RL01830).

[18] B.M. Jenkins, L.L. Baxter, T.R. Miles Jr., T.R. Miles, Combustion properties of biomass,Fuel Processing Technology 54 (1998) 17–46.

[19] S. Van Loo, J. Koppejan, The Handbook of Biomass Combustion and Co-Firing,Earthscan, 2008.

[20] C.R. Woodcock, J.S. Mason, Bulk Solids Handling: An Introduction to the Practice andTechnology, Chapman & Hall, 1996.

40 J.M. Craven et al. / Fuel Processing Technology 119 (2014) 32–40

[21] P.A. O'neill, Industrial Compressors: Theory and Equipment, Butterworth-Heinemann,Oxford, 1993.

[22] D. Schell, High pressure solids feeding using a lockhopper system: designand operating experience, Applied Biochemistry and Biotechnology 17 (1988)73–87.

[23] F.S. Lau, D.A. Bowen, R. Dihu, S. Doong, E.E. Hughes, R. Remick, R. Slimane, S.Q. Turn,R. Zabransky, Techno-economic Analysis of Hydrogen Production by Gasification ofBiomass, US-DOE, NREL, 2002. (DE-FC36-01GO11089).

[24] T.M. Grace, E.W. Malcolm, M.J. Kocurek, Pulp and Paper Manufacture — Volume 5:Alkaline Pulping, 1989.

[25] A.W. JENIKE, Storage and Flow of Solids: Bulletin No. 123, 53Bulletin of the Universityof Utah, 1964.

[26] D. Neves, H. Thunman, A. Matos, L. Tarelho, A. Gómez-Barea, Characterization andprediction of biomass pyrolysis products, Progress in Energy and Combustion Science37 (2011) 611–630.

[27] D. Peralta-Solorio, Personal Communication, E.ON, 2013.