1.37666!CHRISGAS_D130_Mass and Energy Balance Model for Whole Plant

8/13/2019 1.37666!CHRISGAS_D130_Mass and Energy Balance Model for Whole Plant

1/32

Clean Hydrogen-rich Synthesis GasContract No: SES6-CT-2004-502587

Report No. CHRISGAS_December 2009_WP15_D130

Deliverable D130

Mass and Energy Balance for the Whole Plantand Suggestions for Optimisation

Authors: G. H. Huisman, J. Brinkert, G.L.M.A. van Rens, R.L. Cornelissen

Work Package:WP15,Task:WP15.1 2

Contributing &Responsible Partner: CCS

Distribution: ConfidentialDate: 31 December 2009

Revision history:

Rev. no. Date Change information

0


2/32

CHRISGAS_December 2009_WP15_D130:Mass and Energy Balance for the Whole Plant and Suggestions for Optimisation

2(32)


3/32


3(32)

Table of Contents

1 Introduction...........................................................................................................................4

2

Plant description and mass balances ..................................................................................5

2.1 General information on the plant..............................................................................52.2 Present-day case ...........................................................................................................7

2.2.1 Biomass fuel preparation and drying....................................................................72.2.2 Gasification ..............................................................................................................72.2.3 Syngas preparation ..................................................................................................82.2.4 Acid Gas Removal (AGR) .................................................................................. 102.2.5 Fuel synthesis ........................................................................................................ 11

2.3 Near-future case ........................................................................................................ 132.3.1 Methanol................................................................................................................ 132.3.2 Dimethyl Ether (DME)....................................................................................... 13

2.3.3 Hydrogen............................................................................................................... 132.4 Composition of Streams .......................................................................................... 14

2.4.1 Methanol................................................................................................................ 142.4.2 Dimethyl Ether (DME)....................................................................................... 152.4.3 Hydrogen............................................................................................................... 15

2.5 Overall Mass Balance ............................................................................................... 162.5.1 Methanol Mass Balance ....................................................................................... 162.5.2 Dimethyl Ether (DME) Mass Balance .............................................................. 172.5.3 Hydrogen Mass Balance ...................................................................................... 17

3 Energy flows and efficiencies........................................................................................... 18

3.1 Definitions used in energy and efficiency calculations........................................ 18

3.2 Power consumption.................................................................................................. 183.3 Energy flows.............................................................................................................. 19

3.3.1 Methanol Overall Energy Balance ..................................................................... 193.3.2 Dimethyl Ether (DME) Overall Energy Balance ............................................ 213.3.3 Hydrogen Overall Energy Balance .................................................................... 22

3.4 Efficiencies................................................................................................................. 223.4.1 Biomass-to-Fuel Chemical Conversion efficiency........................................... 223.4.2 Energy efficiency of the biomass-to-fuel conversion plant ........................... 233.4.3 Equivalent biomass-to-fuel conversion efficiency....................................... 243.4.4 Exergetic Efficiency............................................................................................. 25

4 Suggestions for Optimisation........................................................................................... 27

4.1 Sale of Process Gases ............................................................................................... 274.1.1 General idea .......................................................................................................... 274.1.2 Implications........................................................................................................... 27

4.2 Use of externally produced (liquid) oxygen .......................................................... 274.3 Production of lower grade methanol ..................................................................... 284.4 AGR for hydrogen plant.......................................................................................... 284.5 Recycle of purge gas in hydrogen plant................................................................. 284.6 Combined Cycle Plant in Hydrogen Plant ............................................................ 294.7 Production of liquid hydrogen................................................................................ 294.8 Direct synthesis of DME from syngas .................................................................. 29

5 Conclusions and recommendations ................................................................................ 316 References ........................................................................................................................... 32


4/32


4(32)

1 IntroductionWithin the framework of European research project CHRISGAS, several plants for theproduction of motor fuels from biomass are designed in a desktop study. Methanol,DME and hydrogen were produced. The investigation is based on a gasification and gasprocessing plant with associated fuel production with a biomass input of 40.8 metric tonper hour dry wood. This work is part of Work Package 15.

In this report mass and energy balances of methanol, DME (dehydrated methanol) andhydrogen production from biomass are presented. The efficiency of the production ofthese fuels is studied. Options for process optimization are identified.

This work is based on prior work on the synthesis and/or production of methanol,DME and hydrogen fuels as described in [Huisman et al., 2009a] and [De Lathouder et

al., 2009]. The present report will be the base for the cost estimation as described in[Huisman et al., 2009b].


5/32


5(32)

2 Plant description and mass balances2 .1 Ge n e r a l i n f o r m a t i o n o n t h e p l a n tFor further discussion the plant is divided in four main sections:

Wood preparation and drying Wood gasification Syngas preparation (reforming, particle removal and shift reactor) Fuel synthesis

A broad outline is given below. Some details are discussed in following sections. Inseveral stages of the process heat is removed. All selected routes have the first threesections in common with sometimes minor changes in process conditions. Theintegrated process has been modelled with a flowsheeting program. However, a smallnumber of devices were placed outside the scope of the model, being the woodpreparation, drying and the air separation unit (i.e. pure oxygen is added to the gasifier).

The process conditions and layout of the plant have been based on careful analysis of theavailable options and their advantages and disadvantages.

Two different plants were designed to allow for the influence of new developments. Oneplant that is representative for the case if one would like to build a biomass-to-fuel planttoday. This means that it is based on technology that is available today, although notalways used at similar conditions. The process conditions are chosen not to be far from

what is considered reasonable now. This case will be called the present-day case, it isschematically shown in Figure 1. The second plant is representative for a plant that will

be build in a few years, after more experience is gained, and a number of optimizationsand a number of new developments have come to the market. This case will be called thenear-future case. It is shown in Figure 2.

The main differences between the near-future case and the present-day case are listedbelow:

Piston feeder1of fuel to the gasifier instead of a lock hopper. This will result inreduced consumption of inert gas and higher quality of the gas (less dilution)

Hot gas filter immediately downstream of the gasifier, so without any coolingbetween gasifier and hot gas filter. The gasifier exit temperature is 900C insteadof 850C. More energy is retained in the gas instead of being transferred to the

steam system. Although the latter is not necessarily a loss the prime objective isto produce fuels from biomass with high efficiency

Catalytic reformer at lower temperature (exit gasifier) instead of steam reformerat 1200C. This will reduce the amount of oxygen needed and increase the energycontent of the gas. Less energy is transferred to the steam system.

1The power consumption of such a piston feeder may be quite high. According to a paper of TKE theinstalled power for a 4 ton/hour feeder is 350 kWe. For the current size of about 40 dry ton/hour this maybecome 3.5 MWe [Koch, 2006]. A piston feeder for up to 200 kg/hr biomass has been build within WorkPackage 7 of CHRISGAS


6/32


6(32)

Figure 1. Present-day case (simplified) Figure 2. Near-future case (simplified)

The plant is designed for the green field so no existing infrastructure needs to beavailable. However, it should be possible to connect the plant to the usual utilities likepower and water. For hydrogen the situation is a bit different. Green field conditionsapply, but the location needs to be near a user of the produced hydrogen. The hydrogenis produced in the gaseous state and transportation either by pipeline or by truck, or

liquefaction is not included in the mass and energy balances.


7/32


7(32)

2 .2 P r e se n t - d a y c a se2.2.1Biomass fuel preparation and dryingIt is assumed that the plant is built in a Nordic country. Therefore, wet wood according

to the Northern mix is used. Moisture content is 50%. The biomass is dried with (ifpossible) low temperature waste heat. The wood is dried to a (low) moisture content of10% in order to improve the efficiency of the process. A storage with five day capacity of

wet wood (9.8 kton) is provided to overcome weekend and holidays. Tramp materialslike stones and metals are detected and removed. It is assumed that the amount of trampmaterial is negligible compared to the biomass feed. The wet wood is screened andoversized wood is chipped before entering the dryer. Dried wood is stored in silos withcapacity of three days (3.2 kton).

Table 1. Mass balance biomass preparation and drying

Wet fuel entering the dryer kg/hour 81,540

Dried wood (10% moisturecontent) leaving the dryer

kg/hour 45,300

Water evaporation kg/hour 36,240

2.2.2GasificationThe gasifier is of the circulating fluidized bed type. The gasification agent is an oxygenand steam mixture. The oxygen is 99.5% pure and is produced on site with an AirSeparation Plant (ASU). Steam (25 bar, 400C) is generated on various locations in theprocess. The addition of steam is about 35% of the dry wood fuel which is in line with

previous calculations made by TPS [TPS, 2009]. The amount of oxygen depends on theselected gasifier temperature, as the gasifier temperature is reached by partially burningthe biomass. For the present-day case the gasification temperature was set at 850C. Thepressure of the gasifier was selected at 20 bar(a).

Inside the gasifier the components in the wood are rearranged in other species. Incombination with the oxygen and steam mostly gases and to some extent heavier,possibly condensable, components (tars, benzene) are formed. The gasification process iscomplex and depends on the type of gasifier, geometry, temperature, pressure, fuelproperties and more. Since the current objective is to model the entire process includingsynthesis of the fuels, it is less relevant to design a more or less exact model of thegasification process based on the reaction chemistry (kinetics) and dimensions of the

gasifier. More or less all changes in gas composition due to various choices wherestraightened out in the reformer. However, in order to describe the most realistic processand gas compositions an attempt was made to model the gasification process. Thecalculations made earlier by TPS have provided a base for this [TPS, 2009]. Note that inthe present study carbon dioxide was used as inert gas instead of nitrogen.

The gasifier model comprises of two stages. First the solid wood is decomposed into theseparate elements like carbon, hydrogen, oxygen, water vapour etc and ash. Some of thecarbon reacts with hydrogen to form hydrocarbons like methane (17%), ethylene (6.5%),BTX (3.4%, as benzene) and tars (5.6%, modelled as naphthalene). The gasifier is thenmodelled as an Gibbs equilibrium reactor, with the above mentioned components treated

as inert gases. Also 4% of the carbon is treated as inert in order to simulate the loss of


8/32


8(32)

unreacted carbon. The heat loss2is assumed to be 0.3% of the energy input. The amountof inert gas for the feeding system was based on a lock hopper system. Recycled carbondioxide from the sour gas wash is used for this purpose. The amount of sand used tomaintain the inventory of inert bed material was related to the amount of ash in the

wood. The resulting gas is presented in Table 7 and resembles closely the composition ascalculated by [TPS, 2009] and described in the literature for pressurised oxygen blownsteam gasification.

Table 2. Mass streams in and out of the gasifier (in Mg/hour)

MeOH DME Hydrogen

Gasifier In

Wood (10% m.c.) enteringthe gasifier

45.3 45.3 45.3

Superheated steam 14.2 14.2 14.2

Oxygen 13.7 13.7 13.4

Inert gas (CO2or N2)3 10.8 10.8 6.8

Sand (SiO2) 4.9 4.9 4.9

Gasifier Out

Raw syngas leaving gasifier 82.6 82.6 78.4

Ash leaving gasifier 6.3 6.3 6.3

2.2.3Syngas preparationThere are four major operations in this section. First gas has to be cleaned from particlesin order to prevent fouling and blocking of the reformer. Currently high temperatureparticle filtration is limited to a temperature of about 600C so the gas has to be cooledprior to particle removal. The removed heat is used to generate saturated steam of 25 barpressure.

For the filter a system with ceramic or sintered metal candles seems to be appropriate.The filter candles have to be cleaned regularly and for this purpose part of the purge gasfrom the synthesis section is used after heating. Energy loss by the combustiblecomponents in this purge gas like hydrogen and methane, can be prevented this way. Theremoval efficiency of the high temperature candle filter is about 99.5% for all solid

components.First the hydrocarbons are treated with steam and oxygen to produce CO and H 2. Thissteam reformer mainly converts methane, but converts higher hydrocarbons and some ofthe carbon as well.

224 3HCOOHCH ++

In order to sustain this endothermic reaction part of the gas is oxidised.

2The value used has been calculated assuming reasonable dimensions typical for the capacity of the gasifier

and heat and transfer to the environment with temperature difference of approx. 50C3Nitrogen is for the DME synthesis


9/32


9(32)

In the model it is assumed that about 25% of the methane is inert and remains in thetreated gas as well as 28% of the carbon. Heat loss to the environment is approx. 0.4%.For the present-day case a thermal reformer at 1200C has been selected. At thattemperature most of the methane and heavier hydrocarbons will be converted.

Table 3. Mass balance over the high temperature cooler and particle filter (Mg/hour)

MeOH DME Hydrogen

HT-cooler in

Ash leaving gasifier 6.3 6.3 6.3

Cleaning gas for filter 2.4 2.4 2.4

Raw syngas leaving gasifier 82.6 82.6 78.4

Particle filter out

Syngas inlet reformer 84.9 84.9 80.8

Ash removed from gas 6.3 6.3 6.3

Table 4. Mass balance over the thermal reformer (Mg/hour)

MeOH DME Hydrogen

Thermal reformer in

Syngas inlet reformer 84.9 84.9 80.8

Superheated steam 2.9 2.9 3.0

Oxygen 15.1 15.1 13.8Thermal reformer out

Syngas leaving reformer 102.9 102.9 97.4

The second major operation is the partial conversion of CO to hydrogen in the hightemperature shift reactor.

222 HCOOHCO ++

Before entering the shift reactor the gas has be cooled to the optimum inlet temperature

of about 350C and the water to CO ratio has to be adjusted to 2.0 with injection ofsaturated steam of 25 bar. Cooling takes place in a second high temperature coolerimmediately downstream of the reformer where saturated steam of 25 bar is generated.

After that a steam superheater reduces the temperature to the final value. Saturated steamfrom both high temperature coolers and from coolers elsewhere in the system issuperheated to 400C. The main emphasis of the shift reactor is to convert so much COto hydrogen that the resulting ratio of CO to hydrogen in the syngas at the inlet of thesynthesis section (i.e. make up compressor) has the correct value.

More precisely: =+

2

22

COCO

COH2.0

Therefore part of the syngas can bypass the shift reactor, the remainder is mixed withsaturated steam and introduced in the reactor. The reactor is modelled as a Gibbs free


10/32


10(32)

energy reactor with modest heat loss of 0.4%. Because of the reaction the temperatureincreases to 467C. The gas is cooled to 272C in the shift gas cooler again to generatesaturated steam of 25 bar. After combination with the bypass flow the gas is cooled to152C and the heat is used to preheat the feedwater to boiling temperature in an

economiser. There is still some heat left in the gas which is used in the wood dryingsystem. The last cooler in this section is actually a desaturator. The gas is cooled to 35Cand water condenses and is removed from the system.

Table 5. Mass balance over high temperature shift and economiser [Mg/hour]

MeOH DME Hydrogen

Syngas bypass of HT-shift 34.1 34.1 0

HT-Shift in

Saturated steam addition 6.0 6.0 7.3

Syngas HT-Shift in 68.7 68.7 97.4

HT-Shift out

Gas shift out 74.7 74.7 104.7

Economiser in

Total gas to economiser 108.8 108.8 104.7

Economiser out

Condensate removal 26.3 26.3 22.1

Gas economiser out 82.6 82.6 82.6

2.2.4Acid Gas Removal (AGR)In a usual synthesis process acid gases have to be removed for some reasons. Firstly thesulphur components like H2S, COS and metal sulphides deactivate and poison mostcatalysts. Indeed this is true for the synthesis of methanol where a concentration as lowas 0.1 ppm(v) is required to maintain proper operation of the catalyst [Rep et al., 2008].Secondly the abundant presence of carbon dioxide requires removal for mostdownstream processing.

Fortunately the sulphur content of biomass feedstock is low (approx. 0.04%). Thereforea raw gas (or sour gas) shift as described in paragraph 2.2.3 should present no problems.

The H2S content (120 ppm) is close to what is accepted as the limit for usual (sweet gas)shift reactors (100 ppm). COS and metal sulfides have not been modelled but is expectedthat COS will be hydrolysed in the shift reactor and the presence of metal sulphides is in

very low concentrations. The guard reactor immediately upstream of the methanolreactor will protect the main catalyst from deteriorating.

However, it is important to remove sulphur to as low as reasonable possible levels inorder to maintain long operating times before replacing the catalyst in the guard reactor.

The AGR has not been modelled in detail. A separator was used to remove CO2and H2Sfrom the inlet syngas. The waste (sour) gas has a H 2S content of 460 ppm and is (aftercompression to 27 bar) partly used as purge gas for the wood fuel feed system. It is notclear whether this can be vented to the atmosphere or that environmental laws prohibitthis. The gas has sufficiently high CO2concentration to make CCS (Carbon Capture andStorage) an attractive alternative. However, this has not been investigated.


11/32


11(32)

For the production of hydrogen the AGR is not necessarily needed as is argued insection 2.2.5.3. Therefore the mass balance over the acid gas removal has not been madein Table 6. A fraction of the removed sour gas was recycled to the biomass feed system.

This fraction is 10.8 Mg/hour.

Table 6. Mass balance over acid gas removal section (Mg/hour)

MeOH DME

Acid gas removal in

Gas to acid gas removal 82.6 82.6

Acid gas removal out

Removed sour gas 58.9 58.9

Sweet gas to fuel synthesis 23.7 23.7

2.2.5Fuel synthesisFuel synthesis has been described in more detail in previous reports within WorkPackage 14 of CHRISGAS ([De Lathouder et al., 2009] and [Huisman et al., 2009a]).Only a couple of highlights will be included in the current report to refresh the memory.

2.2.5.1 MethanolThe syngas is compressed to 67 bar and used as make up in the synthesis cycle.Hydrogen, carbon monixide and carbon dioxide react to form methanol.

OHCHHCO

OHOHCHHCO

32

2322

2

3

+

++

These reactions are both exothermic and therefore the methanol reactor can generatesaturated steam of 25 bar. At the exit of the reactor the reactions are close to equilibrium

which means that there is still unconverted CO and H2present. These are recycled backto the reactor after pressurisation. A small amount of gas is purged from the loop inorder to control the build up of inert gases. The steam pressure in the plant was selectedbecause of reasonable temperature difference between the highest temperature in themethanol reactor and the cooling medium (steam).

Purge gas from two locations is used as cleaning gas for the HT particle filter.

Energy from the condensor of the refining column in the purification section of rawmethanol is used in the wood drying section.

A total of 18,237 kg/hour of grade AA methanol is produced.

2.2.5.2 Dimethyl Ether (DME)It is possible to directly synthesise DME from syngas. In recent years a number ofprocesses has been developed. See for instance [Huisman et al., 2009a]. Using theseprocesses would probably increase efficiency although this has not yet been investigated.Since these process are in the early development stage with only relatively small scaledemonstration plant(s) being tested it was decided to use the conventional route.

Methanol is used as feedstock. It is dehydrated to DEM using a catalytic reactor.


12/32


12(32)

OHOCHCHOHCH23332 +

The raw DME is purified to almost 100% DME which is actually better than requiredfor transportation fuel. The flowsheet up to the DME synthesis is identical to the one formethanol synthesis.

The amount of available methanol from the previous paragraph is sufficient to produce13,085 kg/hour DME.

2.2.5.3 HydrogenHydrogen is produced from the syngas by using a PSA, Pressure Swing Adsorption. Theprocess model for the PSA is relatively simple, a separator. About 85% of the hydrogenis recovered in the product as well as 0.05% of the N2, CO and CO2. All other gasesremain in the purge.

The flowsheet for the syngas production had to be changed for a couple of reasons.

Firstly, the hydrogen content of the purge gas is about 11%. This makes it unsuitable forblanketing gas in the feed system. As an alternative nitrogen was chosen that is availablein large quantities from the air separation unit. The amount of inert gas for the lockhopper feed system is 6 mole per kg dry feed4. This applies both for CO2 and fornitrogen.

A second modification was the removal of the bypass for the shift reactor. The objectiveis to produce as much hydrogen as possible and therefore all the CO should have theopportunity to be shifted to hydrogen. Usually a second reactor is added which performsthe same reaction at low temperature (LT-Shift). However, from initial calculations itappeared that the additional conversion of CO to hydrogen by adding this second reactor

was very small. The investment cost would probably not be paid back by the increasedproduction of hydrogen and therefore the LT-Shift was removed.

A last modification concerns the acid gas removal, AGR. For a good operation of thePSA it is not necessary to remove acid gases. Therefore no AGR is required. However,there may be some advantage of removing in particular CO2and reduce the volume flowto the PSA. This depends on the relative cost of both units. This option was notinvestigated further. Before entering the PSA the syngas is pressurised to 25 bar.

The amount of wood available (45,300 kg/hour with 10% d.s.) is sufficient for theproduction of 2,581 kg/hour hydrogen with 99.95% purity5. There is also a purge gasflow (77,639 kg/hour) with considerable amount of energy (10 vol% hydrogen). Withinthe process there is no need for this energy (besides a small amount for cleaning the HT

gas filter) but it can be used to generate electricity or district heat.

4According to Olle Wennberg (SEP). This is the amount that actually becomes part of the syngas. Theamount needed is 10 mol per kg dry feed of which 50% escapes to the environment + 1 mol per kg dryfeed for various purposes (like purging instruments)

5If a very high purity is not desired then it is possible to increase the recovery from 85% to 90% byselecting a lower purity of 99.9% (Hans de Lathouder, CCS)


13/32


13(32)

2 .3 N e a r - f u t u r e c a s eThe modifications made for the near-future scenario concern only the gasification andgas preparation flowsheet. The major result will be a shift of relatively low grade energyin the steam system to chemical energy in the fuels. With respect to increased fuel outputthe results may be spectacular but one should keep in mind that the downside is that lessenergy is available as power and district heat. These consequences will be presented inmore detail in chapter 3.

2.3.1MethanolThe purge gas for the plug flow feeder is about 20% of that for the lock hopper system.The flow (of recycled sour gas, mainly CO2) is reduced to 3,588 kg/hour. Because of thehigher gasifier temperature (900C) the consumption of oxygen should be slightly higherbut apparently the reduced flow of purge gas compensates for this. The flow of oxygen

to the gasifier has not changed significantly.

There is no cooling of the gas between gasifier and catalytic reformer. The change intemperature from 1200C to 900C has resulted in almost halving the oxygenconsumption, from 15,014 kg/hour to 7,931 kg/hour. The catalytic reformer is modelledas a thermal reformer with reactor temperature 900C and approach of +300C. Thismeans that the equilibrium performance is the same as for the thermal reformer at1200C.

In the HT-Shift reactor the steam consumption is increased from 5,956 kg/hour to10,617 kg/hour due to the increased amount of hydrogen.

No technological improvements are foreseen for the synthesis of methanol. The cleanedgas at the inlet of the methanol synthesis has increased by about 18%. This means thatthe demand for heat in the purification section is higher as well. So, although there is lessenergy available, more is required and therefore less is available for generating power andsupplying district heat.

The overall result on the production of methanol is an increase from 18,237 kg/hour to22,765 kg/hour. An improvement of about 25%.

2.3.2Dimethyl Ether (DME)The changes are the same as for the methanol flowsheet. The production of DME is

directly proportional to the feedstock, i.e. methanol, and this means that the productionof DME increases by 25% as well to 16,333 kg/hour. A higher methanol feedstockmeans that more energy is required in the DME synthesis section. In future it may bepossible to synthesise DME directly from syngas. This has not been considered in thenear-future case but is indicated in the chapter on process optimization (chapter 4), as aninteresting option.

2.3.3HydrogenThe consumption of inert gas (nitrogen) is reduced from 6,849 kg/hour to 2,283kg/hour due to the introduction of the plug feeder. There is a slight (2.5%) increase offlow of oxygen to the gasifier. The other changes are similar to those described in section

2.3.1. The results are similar as well. Decrease of oxygen flow to the reformer (from13,735 kg/hour to 7,136 kg/hour) and an increase in saturated steam consumption in the


14/32


14(32)

HT Shift reactor (from 7,289 to 17,329 kg/hour). The final result is an increase ofhydrogen production from 2,581 kg/hour to 3,155 kg/hour (+22%).

2 .4 Com p o s i t i o n o f St r e a m sThe presentation of the composition of all streams may be confusing and is probably notnecessary. Only the most relevant streams will be presented. These are:

Gas leaving the gasifier Gas leaving the reformer and Gas entering the fuel synthesis

2.4.1MethanolComposition of major streams in mole-%.

Table 7 Molar composition of gas out of the listed devices for methanol productionPresent-day case Near-future caseComponent

Gasifier Reformer AGR Gasifier Reformer AGR

CO 10.6 23.6 26.4 11.1 27.4 28.2

CO2 31.1 19.4 3.4 27.5 13.4 2.2

H2 13.5 17.9 64.0 14.1 25.6 64.6

H2O 32.0 36.5 0.6 33.9 30.9 0.5

CH4 9.0 2.0 4.3 9.5 2.0 3.6

C2H4 2.9 0.0 0.0 3.1 0.0 0.0

BTX 0.3 0.0 0.0 0.3 0.0 0.0

Tar 0.3 0.0 0.0 0.3 0.0 0.0

N2 0.2 0.6 1.3 0.3 0.5 1.0

NH3 26 ppm 10 ppm 21 ppm 21 ppm 17 ppm 28 ppm

H2S 197 ppm 137 ppm 14 ppm 177 ppm 122 ppm 10 ppm

HCl 24 ppm 29 ppm 59 ppm 26 ppm 30 ppm 50 ppm

The H2S content in the syngas is higher than probably can be achieved with a properdesigned AGR. However, for the moment 95% removal of all acid gases was assumed.

A couple of other gaseous stream are of interest. These are the purge from the methanolsynthesis loop and the gas from the crude methanol flash tank. Both contain combustiblecomponents and should be recycled back to the system. The purge gas has the highestenergy contains. The hydrogen content is 54.3% and the methane content 25.5%. Otherminor components are CO, N2and CO2. The flash gas from the crude methanol tank hasa high content of methane (52.1%) some CO (4.2%) and a little methanol (5.8%). Theinert components are nitrogen (9.9%) and CO2(24.5%.

The purge gas has high enough pressure (64 bar) to be used for the filter but the flash gashas to be pressurised to 20 bar before it can be used. This has not been modelled on the


15/32


15(32)

flowsheet. Both streams are recycled back to the filter for about 60%. A full recycle isnot possible because inert components have to leave the system.

A last gaseous flow is the flash gas from the topping column in the methanol purificationsection. This is only a small flow (457 kg/s) with little energy content.

2.4.2Dimethyl Ether (DME)The composition of the main streams upstream of the fuel synthesis are identical to whathas been described in the previous section. The composition of streams in the fuelsynthesis section has been described in [Huisman et al., 2009a].

2.4.3HydrogenComposition of major streams in mole-%.

Table 8 Molar composition of gas out of the listed devices for hydrogen production

Present-day case Near-future caseComponent

Gasifier Reformer HT-shift Gasifier Reformer HT-shift

CO 10.9 21.9 7.4 11.9 26.7 8.5

CO2 23.2 17.2 40.5 24.0 14.0 37.5

H2 13.6 18.1 41.9 13.4 24.1 49.0

H2O 31.9 35.0 0.4 34.6 31.1 0.4

CH4 9.0 1.7 2.2 9.5 1.8 2.1

C2H4 2.9 0.0 0.0 3.1 0.0 0.0BTX 0.3 0.0 0.0 0.3 0.0 0.0

Tar 0.3 0.0 0.0 0.3 0.0 0.0

N2 7.8 6.0 7.4 2.9 2.2 2.5

NH3 91 ppm 26 ppm 31 ppm 42 ppm 24 ppm 27 ppm

H2S 154 ppm 119 ppm 144 ppm 162 ppm 124 ppm 138 ppm

HCl 24 ppm 19 ppm 23 ppm 26 ppm 20 ppm 22 ppm

The hydrogen product is 99.95% pure and contains only a small amount of CO and CO2.In the PSA a purge stream of gas is produced at low pressure (1 bar) that contains aconsiderable amount of hydrogen. This gas is used to some extent for cleaning the HTgas filter (4,200 kg/hour) but has no further use within the process. Its heating value ishigh enough for use as a fuel (approx. 4.8 MJ/Nm3HHV). The main components in thepurge gas are CO2 (63.2%), CO (11.5%), Nitrogen (11.5%), Hydrogen (9.8%) andmethane (3.3%). Since no H2S is being removed it is still present in the purge gas at aconcentration of approx. 130 ppm. The given composition is for the present-day casebut is similar for the near-future case.


16/32


16(32)

2 .5 Ov e r a l l M a s s B a l a n c eThe overall mass balance of the process will be presented for the main streams enteringand leaving the system without the intermediate streams. For steam and inert gas (for thefeed system) only the net flow will be given. Steam is generated in the energy recoverysystem and used for gasifier and reformer. The net input is the amount of make up

water. The wood flow at the inlet of the dryer is at 50% moisture content. The amountof wood in the overall mass balances of sections 2.5.1 to 2.5.3 is wood containing 50%moisture. The actual wood flow into the gasifier is 45,300 kg/hour at 10% moisturecontent.

2.5.1Methanol Mass BalanceTable 9. Overall mass balance of a methanol production plant with a dry biomass input of 40.8 ton/h

Present-day case Near-future CaseIn kg/hr Out kg/hr In kg/hr Out kg/hr

Wood 81,540 Methanol 18,237 Wood 81,540 Methanol 22,765

Oxygen 28,670 Purge gas 1,975 Oxygen 21,690 Purge gas 1,899

Sand 4,886 Ash 6,283 Sand 4,886 Ash 6,283

Water 23,084 Dryer vapour 36,240 Water 27,745 Dryer vapour 36,240

Sour gas 48,107 Sour gas 41,876

Waste water 27,242 Waste water 26,699

Fusel 96 Fusel 96

Total in 138,180 Total out 138,180 Total in 135,861 Total out 135,858

The difference between in and out is very small, as can be expected. The water in theIn column is the make up water for the steam system since superheated steam addedto the gasifier and reformer, and saturated steam added to the HT-Shift are lost. The ashis actually a mix of bedmaterial (sand), ash and carbon. The carbon flow is 546.4 kg/hourand the content in the solid waste is therefore 11.2% (mass). This is the same for bothcases since no modifications are made to the gasifier process6. Most of the waste water isproduced in the desaturator upstream of the acid gas removal. This flow is 27,465

kg/hour for the present-day case or 80% of the total.

6Although the slight difference in temperature of 50C will have some effect on the carbon conversionand actually probably also on the conversion of tars and BTX.


17/32


17(32)

2.5.2Dimethyl Ether (DME) Mass BalanceTable 10. Overall mass balance of a DME production plant with a dry biomass input of 40.8 ton/h

Present-day Case Near-future Case

In kg/hr Out kg/hr In kg/hr Out kg/hr

Wood 81,540 DME 13,085 Wood 81,540 DME 16,333


Sand 4,886 Ash 6,283 Sand 4,886 Ash 6,283

Water 23,084 Dryer vapour 36,240 Water 27,745 Dryer vapour 36,240

Sour gas 48,107 Sour gas 41,876


Fusel 96 Fusel 96


The mass balance for DME is very similar to the one for methanol because a stagedprocess is used and DME is produced from methanol instead of direct synthesis.

The changes when compared to the mass balance for methanol are that the methanol isdehydrated so the flow of fuel (DME) is reduced and the amount of waste waterincreased by the same.

2.5.3Hydrogen Mass BalanceInstead of recycled sour gas nitrogen is used for the fuel feed system. As usual, the purgegas is partly used for the cleaning of the HT gas filter. The net flow of purge gas, whichcan be used as fuel gas, is included in the balance.

Table 11. Overall mass balance of a hydrogen production plant with a dry biomass input of 40.8 ton/h

Near-future case Present-day case

In kg/hr Out kg/hr In kg/hr Out kg/hr

Wood 81,540 Hydrogen 2,581 Wood 81,540 Hydrogen 3,155


Sand 4,886 Ash 6,283 Sand 4,886 Ash 6,283

Nitrogen 6,849 Dryer vapour 36,240 Nitrogen 2,283 Dryer vapour 36,240

Water 24,417 Waste water 22,102 Water 34,456 Waste water 26,107



18/32


18(32)

3 Energy flows and efficiencies3 .1

D e f i n i t i o n s u s e d i n e n e r g y a n d e f f i c i e n c y c a lc u l a t i o n s

The energy flow in and out of the system can be presented in two ways, based on higherheating value (HHV) of the feedstock and the fuels and based on the lower heating value(LHV). In the power plant industry the former is dominantly used in the USA while inEurope efficiencies are usually based on the LHV of fuels. For the actual performance ofa plant the difference is irrelevant, it is obvious that the same amount of steam orelectricity is generated irrespective of the calculation basis for the efficiency.

The difference between HHV and LHV is the energy released when water vapour that isformed from the hydrogen in the fuel and the moisture content is condensed. Forcalculations based on the higher heating value al water is assumed to be in the liquidstate. For biomass fuels the HHV of the dry fuel is approximately 6.9% higher than theLHV.

For a process plant like a wood gasifier followed by synthesis of fuels it is moreconvenient to use the HHV as a basis for the analysis of energy flows. The efficiency ofthe system can be expressed using either of them. Important is always to state which onehas been used HHV or LHV since there usually is a little difference in the numerical

value.

Similar to the approach for the overall mass balance the energy flows to and from thesystem will be presented and not the intermediate values. Energy flows like (district) heatand power are added. Power is included in both the In column and the Out columnto indicate that actual power is generated in the system.

3 .2 Po w e r co n s u m p t i o nTable 12. Power consumption in the total plant

Present-day case (kW) Near-future case (kW)Plant Section

Methanol DME Hydrogen Methanol DME Hydrogen

Wood preparation 100 100 100 100 100 100

Wood drying 850 850 850 850 850 850

ASU 11,468 11468 10,861 8,676 8,676 8,342

Gas preparation 700 700 700 700 700 700

Fuel synthesis

MeOH

DME

Hydrogen

6,100 6,100

150

1,600

7,625

7,625

188

2,000

Total 19,218 19,368 14,111 17,951 18,139 11,992

Not all power consuming devices were included in the model. One of the major

consumers, the air separation unit was not part of the model. In order to get a properoverview of the energy flows, the energy balance as obtained from the modelling


19/32


19(32)

software is altered to include those power consuming devices. In order to keep thebalance balanced, an additional item was added to the out-side of the balance, calledexternal losses.

Table 12 lists the total power demand (in kWe) that has been estimated for each

subsystem based on previous assessments, information from suppliers (ASU), others(dryer by SEP), the calculation flowsheet and reasonable assumptions.

The following additional assumptions have been made:

The power requirement of the ASU has been scaled to actual oxygenconsumption

The power demand of the fuel synthesis for the near-future case is +25% No allowance was made for the expected higher power consumption of the

piston feeder for the near-future case

The total power demand is sufficiently large to consider internal power generation with abiomass combustion plant.

3 .3 En e r g y f l o w s3.3.1Methanol Overall Energy Balance

The following information was used:

HHV of dry wood 20,356 kJ/kg7, dry wood flow 40,770 kg/hour Enthalpy of oxygen at 25C is 23 kJ/kg Enthalpy of sand assumed to be 20 kJ/kg Make up water is assumed to be at 1 bar 25C and enthalpy 105 kJ/kg The power is used for the make up and recycle compressor in the methanol

synthesis plant and two small pumps in the energy recovery

The heat is used to preheat the cleaning gases for the filter HHV of methanol is 22,900 kJ/kg Enthalpy of ash at 600C is assumed to be 680 kJ/kg and 996 kJ/kg at 879C (for

the near-future case)

Carbon loss is 546 kg/hour with heating value 33.8 MJ/kg Energy for dryer vapour comes from LT Cooler and part of the condensor of the

methanol distillation, requirements calculated by SEP

Sour gas as CO2at 35C with enthalpy 29 kJ/kg Waste water at average temperature 35C and enthalpy 147 kJ/kg Fusel, HHV assumed of methanol HHV purge gas methanol synthesis loop 31,600 kJ/kg (combined streams) and

gas leaving the topping column 9,600 kJ/kg (18% methanol)

7In the flowsheet program a value of 19,903 kJ/kg was used in order to compensate for the carbon loss.


20/32


20(32)

Both energy balances close very nicely. It can be noted that in the near-future case thereis too little energy available to generate power. There is also no energy left for districtheat. There is a need for additional heat to supply enough steam for the system. Thiscould be heat supplied by a small wood fired boiler, or a boiler burning the purge gas.

A small modification has been made to the near-future flowsheet that has not beenmentioned before. Because the exit temperature of the reformer is lower than in thepresent-day case the temperature inlet superheater approached the steam temperature of400C. Therefore the inlet temperature of the HT-Shift was raised by 30C. The actualinlet of the shift reactor (after steam addition) is now 345C.

The inlet temperature of the gas of the superheater is now 415C which means that thetemperature difference between gas and steam is small. This leads to a large heatingsurface. However, for the time being it is a reasonable solution. A better alternative canbe to move the superheater to the exit of the catalytic reformer. The combination of hightemperature and corrosive gases is not attractive but maybe in the future suitable

materials are available.Table 13. Overall energy balance of a methanol production plant with a dry biomass input of 40.8ton/h

Present-day case Near-future case

In kW Out kW In kW Out kW

Wood 230,532 Methanol 116,008 Wood 230,532 Methanol 144,811

Oxygen 179 Purge gas 14,567 Oxygen 139 Purge gas 15,189

Sand 27 Ash (hot) 1,187 Sand 27 Ash (hot) 1,738

Water 722 Carbon loss 5,126 Water 809 Carbon loss 5,126

Power 19,218 Dryer vapour 33,300 Power 17,951 Dryer vapour 33,300

Heat 944 Power 9,260 Heat 5,831 Power 0

Sour gas 387 Sour gas 337


Fusel 611 Fusel 611

District Heat 24,767 District Heat 0

Waste Heat 30,664 Waste Heat 40,938

Heat loss 2,400 Heat loss 2,400

Balance 1,097 External loss 13,330 Balance 1,977 External loss 11,726



21/32


21(32)

3.3.2Dimethyl Ether (DME) Overall Energy BalanceAs for the mass balance, the overall energy balance for DME is very similar to methanol.Actually the input is identical except a small increase of the power.

Table 14. Overall energy balance of a DME production plant with a dry biomass input of 40.8 ton/h



Wood 230,532 DME 115,152 Wood 230,532 DME 143,735



Water 722 Carbon loss 5,126 Water 809 Carbon loss 5,126Power 19,368 Dryer vapour 33,300 Power 18,139 Dryer vapour 33,300

Heat 6,423 Power 8,703 Heat 20,278 Power 0



Fusel 611 Fusel 611

District Heat 24,830 District Heat 8,469



Balance 570 External loss 13,458 Balance 1,647 External loss 11,914


For HHV of DME a value of 31,681 kJ/kg is assumed Not enough waste heat available for the dryer, additional input required Additional heat is needed in the DME flowsheet in the destillation section and

for preheating the methanol feed to the reactor

Waste heat comes from the condensors of the two separation columns in thepurification section


22/32


22(32)

3.3.3Hydrogen Overall Energy Balance For HHV of hydrogen a value of 141,567 kJ/kg is assumed Not enough waste heat available for the dryer, additional input required

Table 15. Overall energy balance of a hydrogen production plant with a dry biomass input of 40.8 ton/h



Wood 230,532 Hydrogen 101,445 Wood 230,532 Hydrogen 124,068



Water 712 Carbon loss 5,126 Water 1,005 Carbon loss 5,126

Power 14,111 Dryer vapour 33,300 Power 11,992 Dryer vapour 33,300

Heat 402 Power 9,400 Heat 7,755 Power 2,598


Waste water 902 Waste water 1,066

District Heat 19,725 District Heat 0



Balance 1,814 External loss 12,407 Balance -1,881 External loss 10,198


3 .4 E f f i c i e n c i e s3.4.1Biomass-to-Fuel Chemical Conversion efficiency

A frequently used definition of efficiency is the chemical conversion of energy in thewood to energy in the (synthetic) fuel. These values can be easily calculated from theenergy balance of the various cases.

Table 16. Overview of biomass-to fuel chemical conversion efficiencies (%)



Efficiency HHV 50.3 50.0 44.0 62.8 62.3 53.8

Efficiency LHV 53.9 55.6 45.4 67.3 69.4 55.5

As expected the efficiencies of the near-future cases are higher than those of the present-day cases. The hydrogen case has the lowest biomass-to-fuel conversion efficiencybecause of the large gas stream that is purged. However, it should be noted that in this

definition there is no penalty for power consumption and no bonus for products other


23/32


23(32)

than synthetic fuel (like district heat and energy in the purge gas). Therefore the processis defined in terms of energy efficiency, as is done in the section below.

3.4.2Energy efficiency of the biomass-to-fuel conversion plantThis efficiency is simply the (useful) energy out, divided by the (useful) energy in. Theuseful energy is defined as the sum of energy in biomass, power and heat, suitable fordistrict heating. It is assumed that the energy in the purge gas can be converted to heatsuitable for district heating. The conversion efficiency of the purge gas is assumed to be100% on LHV basis, which is 89% on HHV basis for the methanol and DME case and92% on HHV basis for the hydrogen case. Note that it may be possible to generatepower (and heat) from the purge gas. This is demonstrated in the Vrnamo plant for gas

with a similar heating value as the purge gas of the hydrogen plant, which has the lowestenergy content (expressed in J/Nm3) from all the purge streams. However, this is notconsidered here. Generated power and heat are used as much as possible within theprocess; this means that there is only a net input of electricity and a net output of heatand a biomass stream in and fuel stream out.

Table 17. Energy efficiency of biomass-to-fuel conversion plant (energy in kW, efficiency in %)



Energy in 240,490 241,197 235,243 248,483 248,671 239,926

Biomass 230,532 230,532 230,532 230,532 230,532 230,532

Electricity 9,958 10,665 4,711 17,951 18,139 9,394

Energy out 152,796 146,524 171,561 152,498 145,444 174,307

Fuel 116,008 115,152 101,455 144,811 143,735 124,068

District heat 36,788 31,372 70,116 7,687 1,709 50,239

Efficiency HHV 63.5 60.7 72.9 61.4 58.5 72.7

When comparing the efficiencies listed in Table 16 and Table 17 the striking difference isthe increase in efficiency of hydrogen production relative to the increase in efficiency ofmethanol and DME production. This can solely be attributed to the large stream ofpurge gas, which is converted to district heat. Although hydrogen creation looksinteresting from energy efficiency point of view, it should be noted that this process isthe least efficient when expressed in biomass-to-fuel conversion efficiency and exergeticefficiency. It should be noted that it was implicitly assumed that the generated districtheat is consumed throughout the entire year. It may be worthwhile to co-site with acontinual user. The efficiency for DME production is a little worse than for methanol.

The reason for this is the additional need for heat in the purification section with noincrease in energy output of the fuel.

When expressing the energy efficiency of the plant as defined in this section, one doesnot take into account the energy conversion efficiencies for creating heat and electricity.

Those energy conversion efficiencies are taken into account in section 3.4.3.


24/32


24(32)

3.4.3Equivalent biomass-to-fuel conversion efficiencyIn order to account for the conversion efficiencies of power and heat generation, it wasdecided to express the efficiency in an equivalent biomass-to-fuel conversionefficiency. The required electrical power and the delivered district heat are converted in

an equivalent amount of biomass, needed to create the power and district heat. This isschematically illustrated in Figure 3. It is assumed that biomass can be converted intoelectricity with 30% efficiency (LHV), and into heat with 85% efficiency (LHV). Thisequates to 1.439 kg/hour biomass (50% moisture content for each kWe powerconsumption and 0.508 kg/hour biomass for each kWth heat consumption. Thisequivalent biomass is then added (net energy required by the process) or subtracted (netenergy delivered by the process) from the actual biomass input. This virtual amount ofbiomass is then used to determine the equivalent biomass-to-fuel conversionefficiency. Note that the power and heat plants are virtual plants.

Figure 3. Schematic overview of biomass equivalent conversion

For the calculation of the equivalent biomass-to-power conversion efficiency it doesnot matter, whether net-values of energy are used, or not, as both energy in and energyout are converted using the same conversion efficiency. The equivalent biomass input(converted back to energy content) is then used to calculate the percentage of equivalentbiomass energy converted into fuel.

Table 18. Equivalent biomass-to-fuel conversion efficiency



Net heat in kW -36,788 -31,372 -70,116 -7,687 -1,709 -50,239

Net power in kW 9,958 10,665 4,711 17,951 18,139 9,394Heat equivalentbiomass in

kg/h -18,688 -15,937 -35,619 -3,905 -868 -25,521

Power equivalentbiomass in

kg/h 14,330 15,347 6,779 25,831 26,101 13,519

Biomasscorrection in

kg/h -4,359 -590 -28,840 21,926 25,233 -12,003

Equivalentbiomass in

kg/h 77,181 80,950 52,700 103,466 106,773 69,537

Efficiency HHV % 53.2 50.3 68.1 49.5 47.6 63.1

Efficiency LHV % 57.0 56.0 70.3 53.0 53.0 65.1


25/32


25(32)

It is remarkable that the near-future case actually has the lowest efficiency, although theenergy production in the fuel has increased considerably. This is due to less heat availablefor the steam system with consequently no power generation and less heat for districtheat.

However, the definition used in this section and section 3.4.2 do not take into accountthe quality of the energy available. Especially district heat is a low grade energy. Anefficiency based on exergy will probably provide the best qualification for a conversionprocess.

3.4.4Exergetic EfficiencyIn order to account for the quality of energy that is used within the plant the exergeticefficiency of the process can be defined.

The (simple) exergetic efficiency is defined as:

in

outex

=

With exthe exergetic efficiency, inand out the exergy of the input and output. If weassume that all input and output is at the reference temperature (298K; 25C) andpressure and in steady-state, the exergy of input and output can be defined as follows:

inQininbiomasschemicalin P __, &&& ++=

outQoutoutfuelchemicalout P __, &&& ++=

With & the exergy per unit of time, and P the electrical power per unit of time. Subscript,

chemical refers to the chemical exergy of the fuel or biomass, and Q refers to the heat.The chemical exergy is related to chemical energy by a conversion factor, which is listedin Table 19 for the chemicals under consideration.

Table 19. Conversion factor for energy into exergy (exergy =constant times energy)

Wood 1.05Methanol 0.99DME 0.98H2 0.84

The exergy of the heat streams is defined as follows:

QT

T

Q

ref

Q

= 1&

With Trefthe reference temperature, Tqthe temperature of the heat source into or out ofthe process and Q the exchanged heat per unit of time.

The system boundaries are determined as follows. Generated power and heat are used asmuch as possible within the process; this means that there is, besides biomass and fuelstreams, only a net input of electricity and a net output of heat. It also implies that it isassumed that the required high temperature heat can be generated within the process.


26/32


26(32)

In order to calculate the exergetic efficiency of the biomass-to-fuel plants the referencetemperature is set at 298 K (25C) and the temperature of the heat output is at 363 K(90C).

Table 20. Exergetic efficiency of the biomass to fuel process



Exergy in 252,017 252,724 246,770 260,010 260,198 251,453

Exergy biomass in 242,059 242,059 242,059 242,059 242,059 242,059

Exergy power in 9,958 10,665 4,711 17,951 18,139 9,394

Exergy out 121,435 118,467 101,827 144,739 141,166 118,176

Exergy fuel out 114,848 112,849 89,272 143,363 140,860 109,180

Exergy heat out 6,587 5,618 12,555 1,376 306 8,996

Exergetic efficiency 48% 47% 41% 56% 54% 47%

The exergetic efficiency shows that it is preferable to increase fuel yield, and to reduceheat generation. The near future cases are identified as much more promising. At thesame time the efficiency of hydrogen generation is lower, mainly because of the largestream of purge gas that is converted into district heat. From the viewpoint of exergy thispurge gas, should be recycled into the process. Power generation from purge gases willonly result in a marginally higher exergetic efficiency because of the efficiency of thepower generation process. For the present-day case efficiency would increase from 41%

to 44%, assuming a 36% efficiency of the conversion of purge gas into electricity.


27/32


27(32)

4 Suggestions for Optimisation4 .1

Sa l e o f P r o c e s s Ga s e s

4.1.1General ideaBoth oxygen and nitrogen are produced in the gaseous phase. In the presentconfiguration nitrogen is wasted. However, nitrogen has a monetary value. Pure nitrogencould be sold, when co-siting or liquefied and subsequently sold. The price for liquefiednitrogen in the Netherlands is 100 per ton

For each ton of oxygen the theoretical production of nitrogen is 3.4 ton. This means thatfor say the present-day methanol plant there is potential to supply 95.6 ton per hourliquid nitrogen, or 9,560 per hour, with an annual potential income of 73.7 million(based on 7709 hours of operation per year, see [Huisman et al., 2009b]). It should benoted, however, that part of the nitrogen may evaporate during storage. Location of theproduction plant near one or more large consumers is beneficial.

CO2has a monetary value as well. At the moment the stream coming from the AGR isnot pure, however it should be possible to modify the AGR in order to produce a largehigh grade CO2stream and a smaller stream diluted with H2S. The market price for liquidCO2 depends on location and potential users. A market price of 100 per ton iscommonly used [Janssen et al., 2009]. CO2has many applications and there may well be amarket for Green CO2. It may even be possible to store the CO2 stream, therebycreating a process that actually reduces the global CO2concentration. The potential isabout 300,000 ton per year CO2 assuming that 20% of the CO2 stream is lost. The

market value is 30 million.The total potential for both gases is over 100 million annually8.

4.1.2ImplicationsThe ASU has to be modified in order to supply liquid nitrogen. A large storage for liquidnitrogen is needed and facilities to fill up tanker trucks. The AGR has to be modified (ifnecessary) to produce a large stream of high grade CO2. The gas has to be pressurisedand liquefied. A large storage is required with facilities to fill up tanker trucks.

Alternatively the plant can be located near one or more large consumers and the gas canbe transported using a pipeline network. As a result the capital cost will increase,

probably considerably, but as mentioned the expected revenues are big as well. Theoverall efficiency will probably be lower because of the expected higher powerconsumption.

4 .2 U se o f e x t e r n a l l y p r o d u c e d ( l i q u i d ) o x y g e nInstead of trying to sell the process gases it may be interesting to buy (liquid) oxygen.

The main idea is that since the scale of the present ASU is relatively small, it may bepossible to produce the oxygen on a larger scale at reduced energy consumption cost and

8The market price for methanol is currently (October/December 2009) 223 per ton. The income frommethanol in the present-day case is then 32.5 million, only a third of the potential of the industrialgases.


28/32


28(32)

at reduced cost. Especially when the biomass plant cannot sell the produced nitrogen.The most interesting option would be co-siting with a large producer of pure gases,possibly supplying the producer with the sour gas. Alternatively liquid oxygen could bepurchased.

4 .3 P r o d u c t io n o f l o w e r g r a d e m e t h a n o lThe design of the plant is for the production of high purity grade AA methanol. This isthe main reason why there are two columns in the methanol distillation section, insteadof one. However, the raw methanol entering this section is already pure. The main taskof the second column is to reduce the ethanol content to below 10 ppm. For applicationas a fuel this is usually not relevant. It also may not be relevant for the conversion toDME. It reduces investment costs by about 1 million. It increases process efficiency,because the reboiler of this column is a large consumer of 5 bar saturated steam (approx.20 MWth). Removing the column will reduce the amount of waste heat and more heat isavailable for sale as district heat. Improvement may be up to 2 to 3%.

4 .4 AGR f o r h y d r o g e n p l a n tA PSA does not need prior removal of acid gases. A large investment can be saved.However, the high content of CO2 in the gas adds to the investment cost of the PSA.

Additionally the heating value of the purge gas is relatively low, a higher heating valuemay improve possibilities for further use.

Since it is not necessary to remove all the CO2and in particular, all the H2S, it may bepossible to use a water scrubber. Water is a cheap solvent and CO2dissolves relativelyeasy. If sale of CO2is considered this can also be a AGR with selective solvent. Thesesystems are widely used for upgrading digester gas, a 50/50 mixture of methane and CO2.

Water can subsequently be stripped by air, or, if not possible because of explosion levels,(partly) by the nitrogen from the ASU. The higher heating value of the purge gas fromthe PSA will enable better use of it as a fuel.

The cost will increase for the waterwash and even more for the AGR with selectivesolvent. However, the cost for the PSA and also the power needed to pressurise thesyngas will be reduced. Selecting an AGR may also make the CO2 available for sale.

There will be a small additional power demand and additional pressure drop. The effecton efficiency is expected to be small. It is an interesting option for further research.

4 .5 R ec y c l e o f p u r g e g a s in h y d r o g e n p l a n tThe purge gas contains about 10% hydrogen. Usually this is (partially) recycled back tothe inlet of the PSA. Since the inert levels are high this is not part of the current process.However, in combination with the previous improvement (paragraph 4.4) the inertcontent is decreased and a recycle is possible. The recovery for hydrogen (now 85%) mayincrease to 95% or better. In case the purge gas is recycled back to the HT-Shift it is alsopossible to convert (some of) the remainder of the CO to hydrogen (currently 7.4 % inthe syngas).

An additional compressor will be needed to pressurise the gas from ambient pressure to20 bar. The physical size of the HT-Shift will increase. This will increase the investmentcost significantly. However, the biomass-to-fuel conversion efficiency will increase

significantly.


29/32


29(32)

4 .6 Com b i n e d C y cl e Pl a n t i n H y d r o g e n P l a n tAs mentioned before, the energy in the purge gas is considerable. However, the heatingvalue of the gas is lower that the usual range applicable for gas turbines. This should notbe a problems because in the past decade the application of LCV gas (Low Calorific

Value Gas) in gas turbines has received much attention. In fact the Typhoon gas turbinein the Vrnamo demonstration plant was able to use LCV gas as a fuel. However, it ispreferable to use gas with a higher heating value, because the efficiency will increase andthe range of available gas turbines will be larger. This could be achieved by (partial)removal of CO2, see also section 4.4. Cost will increase considerably with AGR for(partial) removal of CO2and installation of a CHP plant. However, it is hugely beneficialto use the purge gas and it is better, from exergy point of view to generate power than touse it as district heat.

4 .7 P r o d u c t i o n o f l i q u i d h y d r o g e nThe disadvantage of gaseous hydrogen is that it cannot be stored easily. High pressurestorage is probably not an option. This means that the plant needs to be close to a largeuser connected with a pipeline, or that the hydrogen should be liquefied. It could beinteresting to replace the PSA by cryogenic separation, yielding a liquid hydrogen streamimmediately. This will increase hydrogen recovery, probably close to 100% and it willprovide other pure streams like CO2 and possibly a CO stream. The latter may berecycled to the process, or may be used as a fuel.

4 .8 D i r e c t sy n t h e s is o f DME f r o m s y n g a sThe current synthesis is a two stage process, first making methanol and then dehydratingthe methanol to DME. This is the traditional method for making DME.

In the recent past, however, alternative processes have been developed that are capableto produce DME from syngas. Haldor Tops9 is one of them and also the Japanesecompany JFE has developed a process based on a slurry reactor ([Haldor Topsoe, 1996]and [JFE, 2007]).

9This process produces a fuel grade DME with 70-75% DME. The heat demand for the purification willbe decreased.


30/32


30(32)

Figure 4. Process Flow Diagram of 100 ton per day demonstration plant for the direct synthesis ofDME [JFE, 2007]

It can be expected that the process efficiency will increase and heat demand will bereduced, however, this has not been confirmed.

The gas required for the direct synthesis of DME has different CO/H2 ratio than formethanol. This means some (minor) changes to the HT-Shift and possibly differentsteam consumption. The methanol reactor has to be replaced by a DME reactor. The

process requires some refrigeration (depending on manufacturer) so equipment withample capacity is needed.

Because of process intensification the cost is expected to be lower than for the two stageprocess. Various equipment will not be necessary. However, the cost for the process todirectly synthesize DME from syngas may offset the potential savings. Since this processis in the development phase costs are relatively unknown.


31/32


31(32)

5 Conclusions and recommendationsTwo different designs have been made to produces methanol, DME and hydrogen. One

based on the state of the present technology and a case which is expected to be feasiblein the near future. The processes have been modelled in a flow-sheet program and bothcases have closed mass and energy balances. In the present-day case the biomass-to-fuelconversion efficiency ranges between 44% for hydrogen production and 50.3% formethanol production. The efficiency of DME production is only marginally smaller with50.0%. The new developments taken into account in the near-future case give asignificantly larger biomass to fuel conversion efficiency, ranging from 53.8% (hydrogen)to 62.8% (methanol). However, in the near-future case more net power is required. A lotof energy is contained in purge gases. When the energy in the different purge gas streamsis converted into district heat, and the latter is considered useful energy as well, theenergy efficiencies of the present-day cases are in fact higher, because those processes

can generate significantly more heat and power. The present-day hydrogen plant has thehighest energy efficiency with 73%. However, a comparison based on exergy should givethe best insight in which process is using the biomass most efficiently, because it takesthe quality of the energy into account. The near-future cases are most efficient fromexergy point of view. The most efficient of all studied biomass-to-fuel routes was thenear-future plant producing methanol with an exergetic efficiency of 56%.

Several promising optimizations are identified. These can be summarised as optimisingrevenues by sale of process gases, or co-siting with a pure gas manufacturer or byupgrading the purge gas, for either power generation, or a recycle into the process.Cryogenic separation of hydrogen may be interesting as it yields pure streams and

virtually 100% hydrogen recovery.


32/32


6 ReferencesHaldor Topsoe, World International Property Organisation, WO 96/23755 Preparation

of Fuel Grade Dimethyl ether, 8 August 1996

Huisman, G.H., Lathouder, H. de, and Cornelissen, R.L., Clean Hydrogen-rich SynthesisGas, Synthesis System Study, Report No. CHRISGAS_April 2009_WP14 Deliverable,2009a

Huisman, G.H., Brinkert, J., Rens, G.L.M.A. van, and Cornelissen, R.L., CleanHydrogen-rich Synthesis Gas, Cost estimate of a biomass plant with a fuel input of 20 to80 dry tonnes/hr producing different motor fuels, Report No.CHRISGAS_December2009_WP15_D127+D128, 2009b

Janssen, P.P.C.J., Van den Boogaard, R.G.M. and Broeze, J. Haalbaarheidsstudie naar

mogelijkheden Groen Gas op het Nieuw Gemengd Bedrijf Horst aan de Maas,2009JFE, Slurry Phase DME Direct Synthesis Technology-100 ton/day Demonstration Plant,

Operation and Scale up Study, Yotaro Ohno et al. JFE Holdings Inc. To bepresented at Natural Gas Conversion Symposium, 2007

Lathouder, H. de, Huisman, G.H., and Brinkert, J., Clean Hydrogen-rich synthesis gas,Hydrogen Plant Study, Process Design and Cost Estimate for Hydrogen Plant (from50 t/h biomass), Report No: CHRISGAS_July 2009_WP14 Deliverable, 2009

Koch, T., IEA Workshop Chicago October 2006

Rep, M., Cornelissen, R.L., Clevers, S., De Lathouder, H. and Huisman, G.H., Clean

Hydrogen-rich Synthesis Gas, Budgetary Assessment of Post-CHRISGAS TransportationFuel Installation, Report No: CHRISGAS March 2009_WP3_D14, DeliverableNumber D14, 2008

TPS, Personal communication, 2009

1.37666!CHRISGAS_D130_Mass and Energy Balance Model for Whole Plant

Documents