Puigjaner L. (ed.) Syngas from Waste

Подождите немного. Документ загружается.

run at the desired plant conditions and model outputs are compared, while for the

other type of comparisons typically a sensitivity analysis is performed.

Also the accuracy of the model is checked in terms of the net power,

by-products and clean gas characteristics, as shown in Table 4. Measured data

come from ELCOGAS power plant.

The biggest discrepancy is in the estimation of the ASU consumption and in the

amount of sulphur recovery, which is overestimated. It can be deduced that the

underestimation of the LHV of the clean gas is compensated in the CC with a

bigger efﬁciency. Nevertheless, by considering the results as a whole, it can be said

Table 2 Technical data for the IGCC plant model

Gasiﬁcation

Feedstock 2,600 tons day

-1

Crusher ﬁneness 50–60 lm

Feedstock MC 2%

P of lock hoppers 30 bars

gasif

1,400–1,500C

gasif

25 bars

stoich (mole basis) Raw material dependant

O/C

(mass basis) Raw material dependant

T before the WHB 850C

T before gas cleaning 235C

Max. P before CC 3 bars

Combined cycle (CC)

Air

Conditioned by gasiﬁer and gas turbine cycle

requirements

HP/IP/LP 127/35/6.5 bars

HP/IP/LP T before ST 510/520/265C

Compressor pressure ratio 15.7

before combustor 300C

TOT 540C

P of the saturation column 20.8 bars

T of the ﬂue gas 100C

T of the air to air separation unit

(ASU)

130C

Consumables

NaOH (15 mass%) Syngas and cleaning water dependant

(96 mass%) Cleaning water dependant

Natural gas Feedstock dependant

O Syngas and raw material dependant

OH Proportional to the amount of CO

in the shifted gas

Air separation unit (ASU)

Oxygen purity 85% volume basis

Radiation losses 2% adiabatic reactor

15C

0.9309 bars

188 A. D. Bojarski et al.

that in a conceptual stage this modelling approach represents an accurate enough

approximation of an IGCC power plant, with an error in the range of preliminary

design, where error within 15–20% are acceptable [39].

Many different metrics can be deﬁned to prove the validity of the IGCC overall

model checking model tendencies. Some of them are the raw gas composition or

the LHV; others such as cold gas efﬁciency (CGE, deﬁned in Eq. 31 of chapter

‘‘ Modelling Syngas Generation’’), or the Eff

CCapture

(deﬁned in Eq. 28 of chapter

‘‘ Main Puriﬁcation Operations’’ and revised here as well in Eq. 11) previously

deﬁned, are also used. The following paragraphs deﬁne other efﬁciency terms

usually referred.

The energy efﬁciency of the combined cycle (Eff

) is deﬁned as the ratio

between the net power of the CC (considering the energy consumed by the

Table 3 Modelled units in Aspen Plus



Block Aspen Plus model



Feedstock dust

preparation

Dust preparation: crusher

Combustion chamber: Gibbs reactor

Dryer: stoichiometric reactor

Bag ﬁlters: two phase (VL) ﬂash separator

Lock hopper: mixer

Gasiﬁcation and WHB Gasiﬁer: yield and Gibbs reactors

Solid removal: component splitter

WHB: heating units and a compressor for gas recycling

Syngas cleaning Venturi scrubber: RadFrac column and two-phase (VL) ﬂash separator

Sour water stripper: RadFrac columns

COS hydrolyser: stoichiometric reactor

MDEA absorber: RadFrac columns

Claus plant: stoichiometric and Gibbs reactors and two-phase (VL) ﬂash

separators

WGS reactor: two equilibrium reactors

Rectisol process: RadFrac columns

PSA unit: Component splitter

Liquefactor: ﬂash separator and compressor

CC and HRSG Turbines, compressors and heating units Combustor: Gibbs reactor

Saturator: two-phase (VL) ﬂash separator

HRSG: heating units

ASU Compressors and heating units, and component splitter

Table 4 Overall IGCC plant

validation data

Measured Model Error %

Raw gas ﬂow (kg h

-1

) 202,500 192,272 -5.05

LHV clean gas (kJ kg

-1

) 9,911 9,547 -3.68

Gross power (MW) 308 315 2.27

ASU consumption (MW) 29 27.4 -5.52

Net power (MW) 283 285 0.71

Sulfur (kg h

-1

) 3,113 3,318 6.59

Modelling Superstructure for Conceptual Design 189

compressor), and the total chemical energy contained in the clean gas (cg) which is

fed to the gas turbine, see Eq. 6. Even if the most usual stream is clean gas,

hydrogen or PSA purge gas can also be fed to the GT. Note that even if Eq. 6

refers it as cg, depending on the scenario characteristics, the denominator can have

either of the abovementioned natures.

Eff

Power

net

LHV

ð6Þ

By considering the plant as a whole, the total proﬁtable energy that goes into

the system is related with the ﬁnal electricity to the grid. This energy efﬁciency

(Eff

global

) is calculated as in Eq. 7.

Eff

global

Power

net

feed

LHV

feed

ð7Þ

In an analogous way, in the case of H

production, the global efﬁciency is

calculated as in Eq. 8.

Eff

global

LHV

feed

LHV

feed

ð8Þ

Equation 31 of chapter ‘‘Modelling Syngas Generation’’ and (6) evaluate the

performance of the gasiﬁer and of the CC separately, while the Eff

global

provides a

metric of the overall installation. In order to relate the former deﬁnitions the

efﬁciency associated to the gas cleaning operations (Eff

) can be deﬁned as in

Eq. 9, where m

and m

refer to the raw and clean gas mass ﬂows.

Eff

LHV

ð9Þ

Consequently, Eff

global

can be calculated from the partial efﬁciencies as in

Eq. 10.

Eff

global

¼ Eff

Eff

CGE ð10Þ

In an analogous way, for the H

production the global efﬁciency can be cal-

culated by considering in this case the partial efﬁciencies, including the efﬁciency

of the carbon capture train (Eff

CCapture

), where this is deﬁned last as in Eq. 12.

Eff

CCapture

LHV

ð11Þ

Eff

global

¼ Eff

CCapture

Eff

CGE ð12Þ

The validation of model trends is done by computing the former metrics, by

means of sensitivity analysis (SA). In every SA the LHV of each stream, raw gas

mole ﬂowrates, emissions, by-products, power and water consumption are con-

sidered and taken directly from model results. While the LHV of fuels is calculated

190 A. D. Bojarski et al.

using Eq. 32 in chapter ‘‘Modelling Syngas Generation’’, the estimation of LHV

and LHV

is performed using the Aspen Plus



property sets (QVALNET) [20].

Differences between each case study lie on raw material composition, and they

consider the oxygen ratios to the gasiﬁer and to the combustion chamber in the CC

with an adjustment of the gasiﬁer temperature (T

gasif

), to remain close to the

1,400C and the TOT, to be as close as possible to 540C.

The sensitivity analysis (SA) performed changes considering the gasiﬁer

feedstocks: the equivalence ratio, ER, is changed in an interval of ±30% its base

case value (i.e., a variation from 0.29 to 0.55), while the steam ratio is varied

between 0.15 and 0.4 for all case studies. To point out that due to the oxygen ratio

value into each case study, the ratio, deﬁned as proportionate to the stoichiometric

oxygen of the mixture, is always below the complete combustion value. Eight

values are considered in both ranges; accordingly 64 points are resulted for each

analysis. The CC fed oxygen ratio is maintained constant.

The SA results are represented in two-dimensional graphs, where the axis

corresponds to the input variables and the output variable changes are plotted using

contour lines. Blue points in the graphs are the SAs results, while the red star point

is the highest value, for the output variable in the SA.

O ratio

gasif

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

800

1000

1200

1400

1600

1800

2000

2200

2400

O ratio

TOT

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

490

500

510

520

530

540

550

O ratio

LHV

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

x 10

Fig. 8 Iso-lines for T

gasif

, TOT, and LHV

in C and J/kg, respectively, for the base case

Modelling Superstructure for Conceptual Design 191

Figure 8 reports the variation observed for T

gasif

, TOT and the LHV

.Itis

clearly seen that in the role as a moderator of the steam in the gasiﬁer higher steam

ratios are followed by lower temperatures, and that more inlet oxygen implies

higher temperatures.

The considered SA leads to gasiﬁcation temperatures in the range of 800–

2,400C, even if the gasiﬁcation zone is in the middle, limited by CH

and sulphur

oxides production. TOT is between 550 and 490C. In the modelled installation

and regarding the HRSG, lower temperatures than 540C are not feasible operating

regions due to the steam temperature requirements. With regard to the LHV

,it

can be seen that the highest values are obtained for the combination of lowest H

and ER.

Syngas compositions (H

, CO, CO

and CH

) are represented in Fig. 9. In the

case of CH

its occurrence is mainly driven for the O

amount coupled with high

gasiﬁer temperatures (see Fig. 8 left, and Fig. 9 bottom right). For the case of H

CO and CO

non-linear shapes are found. In the case of H

, the highest H

production is found for an ER close to 0.4, and for the highest ratio of steam fed,

O ratioH

O ratio

O ratioH

O ratio

Mole H

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

Mole CO

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

Mole CO

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

Mole CH

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

Fig. 9 Iso-lines representing the molar fraction (%) of raw gas components in dry basis for the

base case

192 A. D. Bojarski et al.

which is in clear concordance with Eqs. 22 and 23 in chapter ‘‘Modelling Syngas

Generation’’. In the case of CO

, a clear relationship with H

composition is found

but its highest values are found for the maximum amount of O

and steam. As

expected, the CO behaviour is opposite to the one obtained for CO

Figures 10 and 11 are highly interrelated; ﬁrstly, steam and gas turbines power

production are shown in Fig. 10. The amount of power produced by the gas turbine

reaches a maximum when ER and steam ratio are the lowest. On the contrary,

steam turbine production is maximum at higher oxygen ratios but minimum at

lower steam ratios. In both cases the inﬂuence of O

ratio is more important than

the one of steam ratio.

In the case of CGE, it is found in Fig. 11 (top left) that the highest value

coincides with the one obtained for the maximum gas turbine power production,

and is in concordance with the higher LHV

values found in Fig. 8, given that the

fuel LHV remains constant. Similarly, the Eff

global

shows the same behaviour due

to the fact that most of power comes from GT, and the fact that fuel LHV remains

constant. In the case of the highest value for Eff

, it coincides with the highest

steam turbine power production, therefore with highest gasiﬁcation temperatures

the WHB recovers more heat for steam production.

It is demonstrated here that oxygen ratio has higher inﬂuence in gasiﬁcation

than the steam one. The CGE is the main parameter that inﬂuences in the ﬁnal

global efﬁciency of the plant, since it is the GT that produces more power.

3.5 Results

Let us compare the base case, where all the syngas goes to the CC, with the

extreme cases where all the clean gas is used to produce: (a) pure H

to be sold to

the market, (b) pure H

to be sold to the market with PSA purge gas used in the CC

or (c) H

to be used in a hydrogen turbine in a CC. The main difference between

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

170

180

190

200

210

220

230

O ratio

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

100

105

110

Fig. 10 Iso-lines for gas and steam cycles power production (MW) for the base case

Modelling Superstructure for Conceptual Design 193

the (a) and (b) with the (c) scenarios is the use or not of the PSA unit, since if the

is not going to the market, its required purity is not restrictive. Moreover, in all

cases the Rectisol block is used to separate H

from the remaining gases, which

could be proﬁted by a carbon capture scheme, and it is not proﬁted in the GT.

The main advantage of using H

in GT combustion is the lesser level of

emissions to the atmosphere, which hypothetically will involve only water. In

these cases, we are assuming that the GT conceived for syngas (operating and

main characteristics) can be used for H

. Please note that this requires the HRSG

structure to be deﬁned for generating steam from GT ﬂue gas works similarly

when being fed with steam from H

combustion. See in Table 5 the most

Table 5 Comparison between electricity and H

production

Syngas to CC H

to market H

to market - PSA to CC H

to CC

Eff

global

(%) 42 – 5.4 33.3

Eff

global

(%) – 51 51 –

Gross power MW 315 0 85 273

Net power MW 285 -46 37 226

CGE

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

O ratio

Eff

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

Eff

global

0.3 0.35 0.4 0.45 0.5 0.55

0.15

0.2

0.25

0.3

0.35

0.4

Fig. 11 Iso-lines for unitary and global efﬁciencies (%) for the base case

194 A. D. Bojarski et al.

important parameters reported for each case study: the efﬁciency calculated as in

Eqs. 7 and 8 depending on the ﬁnal output, and the gross and net power for each

situation. Note that even if the compressor in the GT cycle is not working for the

GT itself, it is needed to compress the air that goes to the ASU.

The ﬁrst column corresponds to the base case where all syngas is combusted in

the CC. The other columns correspond to the following scenarios that imply

several changes in the ﬂowsheet layout. In all the scenarios performed, the amount

of air that goes to the GT is regulated with a design speciﬁcation in Aspen Plus



order to adjust TOT to a temperature around the 540C:

• H

to market: In this scenario, the HRSG has been deactivated. We are

assuming that no air for the GT combustor is entering the system, and that no

refrigeration is needed for the turbine. The waste N

from the ASU can be sent

to the feeding system, since it is not needed in the GT system.

• H

to market–PSA to CC: It is similar to the previous one. In this case, an

amount of ‘‘residual’’ power is produced. It checks the validity of this PSA

residual gas as feed for the gas turbine, but it does not justify the presence of a

turbine, as seen in Table 5. The HRSG is again activated, as well as the

refrigeration system of the gas turbine.

• H

to CC: In this scenario, as this differs from the two previous ones, the PSA

unit is not used; thus, the hydrogen produced in the WGS reactors is directly

sent to the CC.

Table 5 shows that the efﬁciency penalty of a combined cycle with syngas or

with H

is around 7%. This is due to the extra power consumption of the carbon

removal train, to obtain the ﬁnal H

stream. On the other hand, the increase of

feedstock that would supply the same amount of power in a CC moved by

hydrogen instead of syngas, according to this data is about 21%.

Analogous to chapter ‘‘Main Puriﬁcation Operations’’, the outputs of the

superstructure (power and H

caloriﬁc value) have been analysed versus the per-

centage of syngas stream that goes to the CC. The electricity consumed by the

carbon capture technology has been considered. In the case of H

production, the

consideration here is the production of pure H

to be sold in the market. It is

interesting to appreciate that at the value of 50%, the caloriﬁc value of the H

is the

same than the gross power produced by the CC, thus deducing that both appli-

cations have similar efﬁciencies. Figure 12 shows the lineal dependencies of the

abovementioned parameters. It is interesting to mention that the production of

hydrogen has an almost constant efﬁciency of 68% in all the scenarios. In the case

of the gross power, two different behaviours have been plotted: in dashed black the

one that considers all waste N

available from the ASU being fed to the combustor

of the GT, and in colour the one considering a fraction of the waste N

stream

being separated according to the same fraction that of syngas. This waste N

stream has a ﬂowrate of 246,000 kg h

-1

and a temperature of 384C at a pressure

of 20 bars, and in general is proﬁted to increase the GT overall ﬂowrate, while the

syngas is fed to the GT at 147C. The efﬁciency of the CC is different in the former

two situations: in the ﬁrst one, the efﬁciency increases while the proportion of

Modelling Superstructure for Conceptual Design 195

syngas to the CC decreases, ranging from 63 to 71% for values of the splitter

between 1 and 0.4. In the second situation the efﬁciency is approximately constant

around the 63%. It is due to the importance of the total ﬂowrate that goes to the

GT, which is crucial for the ﬁnal power produced. See Fig. 13 where the efﬁ-

ciencies comparison is plotted.

In both Figs. 12 and 13, note that the lines related with the CC operation are not

plotted for all the split fractions, given that there is a operative limitation of

reaching a TOT value around 540C, in order to make the GT work properly. It is

found that for some split fraction scenarios it is impossible to obtain a value of

TOT close to its requirement, even with the stoichiometric combustion of the

syngas. This means that there is a certain amount of waste N

that refrigerates the

combustion mixture too much, which can also be interpreted that the combustion

mixture does not have the appropriate ﬂowrate to produce the high temperature

gases adequate for the GT operation. Note that the gross power produced with

-20

140

220

300

380

0 20406080100

% Syngas to CC

Gross Power H2 C Capture

Fig. 12 H

, carbon capture

technology consumption and

gross power variation with

the percentage of syngas that

goes to the CC. In colour,

again the gross power but

with waste N

separation

020406080100

EffCC EffCCN2 Effccapture

% Syngas to CC

Fig. 13 Efﬁciencies for the

CC and the carbon capture

sections. Eff

CCN2

contemplates N

separation

196 A. D. Bojarski et al.

waste N

constant is higher than the one that separates it, even if this last option

allows for an operation with a smaller amount of syngas.

4 Conclusion

This chapter has provided an overall view of the possible ways of handling the

IGCC design problem from a ﬂowsheet optimisation point of view (preliminary

design). Clearly, it has shown the importance of the notion of superstructure,

which is introduced in the ﬁrst section. It provides the necessary ﬂexibility and

versatility in the selection of the best alternative design during the synthesis

procedure. Moreover, different algorithms introduced for tackling the design

optimisation are readily accessed by the superstructure. Model validation is

emphasised by means of plant data comparison and by the use of sensitivity

analysis. Finally, the efﬁciency penalty because of CO

capture methods is dem-

onstrated through four case studies. The limitation in poly-generation associated to

the value of TOT in the ﬂue gas stream is analysed in depth. In this sense the

superstructure and its versatility have been proved and ready for scenarios

optimisation.

References

1. Cameron I (2005) Modelling across the process life cycle: a risk management perspective.

Eur Symp Comput Aided Process Eng 15:3–19

2. ISO15288 (2008) Systems and software engineering—system life cycle processes. Technical

report ISO, Genève, Switzerland

3. ISO14040 (1997) Environmental management—life cycle assessment—principles and

framework. Technical report ISO. Genève, Switzerland

4. Cameron IT, Newell RB (2006) Modeling in the process life cycle. In: Puigjaner L, Heyen G

(eds) Computer aided process and product engineering, Sect. 4.2. Wiley, New York, pp 667–

693

5. Yang Y, Shi L (2000) Integrating environmental impact minimization into conceptual

chemical process design: a process systems engineering review. Comput Chem Eng 24:1409–

1419

6. Rebitzer G, Ekvall T, Frischknecht R, Hunkeler D, Norris G, Rydberg G, Schmidt W, Suh S,

Weidema B, Pennington D (2004) Life cycle assessment Part 1: framework, goal and scope

deﬁnition, inventory analysis, and applications. Environ Int 30:701–720

7. Douglas JA (1985) Hierarchical decision procedure for process synthesis. AlChE J 31:353–

362

8. Biegler L, Grossmann I, Westerberg A (1997) Systematic methods of chemical process

design. Prentice-Hall, Englewood Cliffs

9. Yeomans H, Grossmann IE (1999) A systematic modeling framework of superstructure

optimization in process synthesis. Comput Chem Eng 23:709–731

10. Metz B, Davidson O, Coninck H, Loos M, Meyer L (2005) Special report on carbon dioxide

capture and storage, Chap. 3. IPCC, Switzerland

Modelling Superstructure for Conceptual Design 197

Puigjaner L. (ed.) Syngas from Waste - Emerging Technologies

Подождите немного. Документ загружается.