Energy-ef fi cient v acuum systems V acuum technology is used in petroleum ren- eries to facilitate the distillation of heavy ends at low temperatures, to prevent coking and degradation of products and for other applications. Currently, steam jet ejectors and steam ejector-liquid ring vacuum pump (LRVP) combi- nations are the most common methods for vacuum generation in petroleum reneries. Although steam jet ejectors are very reliable, they are highly inefcient. Due to increasing energy costs and environmental concerns, it is essential to reduce the energy required for vacuum generation. Petroleum reneries discard a lot of waste heat to the environment, which could be used to reduce energy consumption for vacuum gener- ation. There are publications illustrating the merits and limi- tations of vacuum generation methods 1,3,4,6 and chilled/refrig- erated water generation. 2 However, there is a need for an integrated approach covering all aspects of vacuum genera- tion and its energy reduction possibilities. The main objective of this Case analysis of the techniques available to reduce energy consumption in vacuum systems reveals the potential for cost savings C CHANDRA SEKHARA REDDY and S V NAIDUAndhra UniversityG P RANGAIAH National University of Singapore www.digitalrefining.com/article/1000789PTQ Q2 20131 study is to analyse various methods for developing ener- gy-efcient vacuum generation systems in petroleum rener - ies. Three case studies are presented to enhance under- standing in the selection process. Vacuum generation in refineries Steam ejectors and LRVPs are generally used in petroleum reneries. A review of various vacuum generation equipment capacities, operating ranges and efciencies is available. 6 Steam ejectors may have one or more stages in series or a series- parallel combination, with pre- and/or interstage condensers, depending on the level of vacuum required and the utility optimisation and operational exibility sought for various plant loads. Steam ejectors are highly reliable, and the availa- bility of steam in petrol eum reneries makes ejectors the natural choice. However, they are highly inefcient 6 (<10%), mainly due to a lack of moving parts to convert uid velocity to pressure efciently. 4 LRVP most commonly uses water as a seal liquid since it can be separated and reused safely. They are generally more expensive compared to steam ejectors. However, they do not require large heat exchangers to condense the vapour at their outlet, and the operating costs of LRVPs are generally lower than steam jet ejectors. For better operating cost savings, a steam ejector–LRVP combina- tion is sometimes used to replace the last one or two stages of a multistage steam ejector system. Design principles and utility requirements This section presents useful design principles and tools for estimating the utility require- ment for steam jet ejectors and LRVP. Use of pre- and inter- stage condensers can reduce both capital and operating costs for the vacuum system. The vacuum produced is limited by the temperature of the cooling water; the colder the temperature of the cooling water, the lower the vacuum produced. Steam requirement for ejec- tors can be estimated based on the dry air equivalent (DAE) of
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Vacuum technology isused in petroleum ren-eries to facilitate the
distillation of heavy ends at lowtemperatures, to prevent cokingand degradation of productsand for other applications.Currently, steam jet ejectors andsteam ejector-liquid ringvacuum pump (LRVP) combi-nations are the most commonmethods for vacuum generationin petroleum reneries.
Although steam jet ejectors arevery reliable, they are highlyinefcient. Due to increasingenergy costs and environmentalconcerns, it is essential toreduce the energy required forvacuum generation. Petroleumreneries discard a lot of wasteheat to the environment, whichcould be used to reduce energyconsumption for vacuum gener-
ation. There are publicationsillustrating the merits and limi-tations of vacuum generationmethods1,3,4,6 and chilled/refrig-erated water generation.2 However, there is a need for anintegrated approach coveringall aspects of vacuum genera-tion and its energy reductionpossibilities.
The main objective of this
Case analysis of the techniques available to reduce energy consumption invacuum systems reveals the potential for cost savings
C CHANDRA SEKHARA REDDY and S V NAIDU Andhra University
study is to analyse variousmethods for developing ener-gy-efcient vacuum generationsystems in petroleum rener-ies. Three case studies arepresented to enhance under-standing in the selectionprocess.
Vacuum generation in refineriesSteam ejectors and LRVPs aregenerally used in petroleumreneries. A review of various
vacuum generation equipmentcapacities, operating ranges andefciencies is available.6 Steamejectors may have one or morestages in series or a series-parallel combination, with pre-and/or interstage condensers,depending on the level ofvacuum required and the utilityoptimisation and operationalexibility sought for various
plant loads. Steam ejectors arehighly reliable, and the availa-
bility of steam in petroleumreneries makes ejectors thenatural choice. However, theyare highly inefcient6 (<10%),mainly due to a lack of movingparts to convert uid velocity topressure efciently.4
LRVP most commonly useswater as a seal liquid since it
can be separated and reusedsafely. They are generally moreexpensive compared to steamejectors. However, they do notrequire large heat exchangersto condense the vapour at theiroutlet, and the operating costsof LRVPs are generally lowerthan steam jet ejectors. For
better operating cost savings, asteam ejector–LRVP combina-tion is sometimes used toreplace the last one or two
stages of a multistage steamejector system.
Design principles and utilityrequirementsThis section presents usefuldesign principles and tools forestimating the utility require-ment for steam jet ejectors andLRVP. Use of pre- and inter-stage condensers can reduce
both capital and operatingcosts for the vacuum system.The vacuum produced islimited by the temperature ofthe cooling water; the colderthe temperature of the coolingwater, the lower the vacuumproduced.
Steam requirement for ejec-tors can be estimated based onthe dry air equivalent (DAE) of
LRVPs is dictated by thetemperature of seal water.With normal cooling watertemperatures of ~30°C, LRVPsare used to replace steam ejec-tors operating at suctionpressures >150 torr (usually the
last one or two stages of ejec-tors for vacuum distillationcolumns). Power required by aLRVP can be estimated usingequation 8:3
P =
Pa * Vs * ln Pd
Pa
27 000 * ηe (8)
Chilled water generationChilled water can be generatedeconomically using absorptionheat pumps and mechanicalrefrigeration. Absorption heatpumps use waste heat (such aslow-pressure steam or a hotprocess stream) rather thanmechanical/shaft energy foroperation. Lithium bromideabsorption pumps arefrequently used due to theirlower cost and application
range up to the freezing pointof water. Compared tomechanical chillers, absorptionchillers have a low co-efcientof performance (COP).Nonetheless, their operatingcosts can be substantially lower
because they use waste heat,while vapour compressionchillers must be motor driven.At lower electricity prices, a
mechanical chiller can beattractive for chilled watergeneration.7
Optimisation of vacuum systemoperating costA design strategy and proce-dure for minimising theoperating cost of a multistagesteam jet ejector system isdiscussed in this section and
suction gases (including air,water vapour and other gases).As per the HEI (Heat ExchangeInstitute) procedure for calcu-lating DAE8 , water vapour inthe suction gases is consideredseparately and all other gases
(including air) are treated as amixture, in accordance withthis mixture’s molecularweight. HEI has publishedcurves to convert suction gasstreams to DAE using molecu-lar weight and temperatureentrainment ratios. Molecularweight entrainment ratio(MW
c) is dened as the ratio of
the weight of suction gas to theequivalent weight of air.Temperature entrainment ratiois dened as the ratio of theweight of air (or water vapour)at actual suction temperatureto the weight of air (or watervapour) at 21.1°C.
The following equations arederived from HEI curves8 fortemperature entrainment ratios(TC
For M = 60 to 150, MWc = -2 * 10-5 M2+ 0.0077M + 0.9464 (3.2)
Water vapour and othercomponents in the suction gascan be converted to DAE usingthe correction factors fromequations 1, 2, 3.1, 3.2 and thefollowing equation:
DAE of suction WOG
+
Ww
gas or vapour=
TCa * MW
COGTC
W * MW
CWV(4)
The amount of motive steamrequired to compress (fromsuction pressure to dischargepressure) the unit DAE mass of
suction gas/vapour in a steamejector is dened as Ra (kg of
motive steam/kg of DAEequivalent of load gas). Valuesof R
a are available4 as curves,
with suction pressure on theabscissa and discharge pres-sure on the ordinate. For roughestimation of an ejector’s steamconsumption, one can simplyuse the following equation:9
Ra =
>Pd(0.434 -
1.338+ 0.000475 Pa) - 0.187H
Pa Pa
(1.2 - P
v - 10.2)
20 (5)
Steam requirement for ejec-tors can be estimated bymultiplying DAE and R
a
values.
For condenser calculations,involving air and watermixtures, the overall heattransfer coefcient, U W/(m2K), can be estimated usingthe following developed equa-tions.4 For a gas vapourmixture with non-condensablevapour mole percentage, NC,from 1% to 50%:
illustrated with two case stud-ies. There are several ways toreduce the operating costs of avacuum system. Engineersoften nd it difcult to takeoptimisation decisions as mostof the information is vendorspecic. The present work
illustrates the use of simpletechniques to optimise vacuumsystem operating costs easilyand quickly. A design strategyfor optimising a new vacuumsystem is presented in Figure 1.Process simulators such asAspen Hysys can be used to
estimate a condensing temper-ature at which the majority(~90%) of vapour condenses, atvarious stages of the vacuumsystem.
Once a vacuum systemdesign is selected, optimumdischarge pressure and steam
Evaluate the type of gasesto be evacuated
Mostly non-condensablegases?
Pre-condenser usingcooling water
Is cheap source of mediumpressure steam available?
Install multistage steam ejectorsystem with optimum design
Install pre-condenser using chilledwater (generated by mechanical or
consumption for each stage ofthe multistage steam ejectorsystem (see Figure 2) can beobtained by solving the follow-
ing optimisation problem; forexample, using the Solver toolin Microsoft Excel. Denitionsof parameters and variables inthis optimisation problem aregiven in the nomenclature.
The following objective func-tion is to minimise overallsteam consumption for themultistage steam ejectorsystem:
Wv = ∑ni =1
Wvi =∑n
i =1
Pdi (0.434 - 1.338 +
Pai
Pai
0.000475 Pai) - 0.187] (1.2 -
Pv-10.2
)20
(Wa +
Wwi ) (9.1)
MW
CWV
The quantities in the aboveequation are as follows.
Suction pressure for (i+1)thstage:
Pai+1
= Pdi - Dp
i (9.2)
Water vapour ow rate tothe inlet of (i+1)th stage:
Wwi+1
= Wwi - 18 * Pv
i
29 * (Pai+1
- Pvi) (9.3)
Saturation pressure of watervapour corresponding to thevent temperature of the ithstage condenser is a function ofthe ith stage condenser venttemperature:
Pvi = f(Tvo
i) (9.4)
Vent temperature for the ithstage condenser:
Tvo
i= T
wco + TAP
i(9.5)
Mole fraction of watervapour at the ith stagecondenser inlet:
nwi =
Wwi + Wv
i / [
Wa+
Wwi + Wv
i] (9.6)
18 29 18
Mole percent of non-conden-sable gases at the ith stage
condenser:
NCi = (1 - nw
i) * 100 (9.7)
Partial pressure of watervapour at the ith stagecondenser inlet:
Ppwi = Pd
i * nw
i (9.8)
Saturation temperature of
water vapour corresponding toits partial pressure at the ithstage condenser inlet is a func-tion of its partial pressure atthe ith stage condenser inlet:
Tvi= f(Ppw
i) (9.9)
Log mean temperature differ-ence for the ith stagecondenser:
Overall heat transfer coef-cient for the ith stagecondenser, U, is given by equa-
tions 6 and 7.Required area for the ithstage condenser:
Ai=
Li * (Ww
i + Wv
i - Ww
i+1) * 1000 (9.11)
Ui * LMTD
i * 3600
Cooling water required forthe ith stage condenser:
Wcwi=
Li * (Ww
i + Wv
i - Ww
i+1) * 0.239 (9.12)
(Twc
- Twco
)
Decision variables: Pdi for i=1,
2,... n-1Bounds and constraints:Pa
1 < Pd
i < Pd
n and Wv
i > 0
Steam properties in equa-tions 9.4 and 9.9 can becalculated using the Excelspreadsheet, freely available atwww.x-eng.com.
The above optimisation prob-
lem (Equation 9) is applicableto systems involving air andwater only. It ignores vapoursuperheat at the ejector inlet,and vapour sub-cooling andliquid sub-cooling in condenser
system with different approachtemperatures are shown inTable 1. The vent pressurestally well with those given inthe reference,4 with error in therange +14% to -8%. For three-and four-stage steam ejector
systems, steam ow rates esti-mated by the above procedureare greater by up to 25%,whereas, for a two-stage steamejector system, the estimatedsteam ow rate was lower by~20% compared to the reportedvalues.4 The main reason is thedifference in Ra values fromthe reference graphs4 andEquation 9.1, used in the opti-misation procedure. Foraccurate estimation of steamow rates, one can use theoptimised vent pressures andthe reference graphs.4 Coolingwater consumption shown inTable 1 can be reduced furtherwith some increase incondenser area by using aseries ow arrangement, wherewater from the rst interstagecondenser ows through the
other interstage andafter-condensers.
For relatively higher suctionpressures (~40 torr), use of apre-condenser, with chilledwater cooling, can further
area calculations. Also, itassumes simple LMTD withoutany correction factor. Hence,condenser area calculations areapproximate. However, theoptimisation problem inEquation 9 is very useful to
arrive at a preliminary designconcept and/or to verifyvendors’ proposals. For largesystems involving othervapours, the optimisation prob-lem can be solved by includingsimulation data from processsimulators such as AspenHysys, Aspen Plus and Pro/II.
The following case study(case study 1), based on thedetails from the manual opti-misation of a vacuum systempresented by Power,4 illustratesthe effectiveness of the optimi-sation procedure in Equation 9.
Load/suction gas: 45.36 kg/hof air + 97.52 kg/h of watervapour at suction pressure of15 torr and suction tempera-ture of 21.1°C; dischargepressure: 815 torr; motivesteam: 10.6 barg.
Cooling water supply andreturn temperatures are 32.2°Cand 37.8°C. Solutions to theoptimisation problem, obtainedusing Excel’s Solver, for a two-,three- and four-stage ejector
Approach Values from optimisation
temperature Utilities Inter-condensers Total Number at 1st Cooling Vent Condenser heatInter- condenser, MP steam, water, temperature, inlet pressure, Inter- and after- transfer
Case Stages condensers °C kg/h kg/h °C mmHg condenser area, m2 area, m2
reduce the steam requirementand the ejector system’s capitalcost. This is illustrated by thefollowing case study (case
study 2), solved using the opti-misation procedure in Equation9. Different quantities of watervapour and possible use ofchilled water are considered,and the results are summarisedin Table 2.Load gas (Cases 2A and 2B):45.36 kg/h of air + 97.52 kg/hof water vapourLoad gas (Cases 2C and 2D):
45.36 kg/h of air + 453.6 kg/hof water vapourSuction pressure and tempera-ture: 40 torr and 33.3°CMotive steam: 10.6 bargCooling water supply andreturn temperatures: 32.2°Cand 37.8°CChilled water supply andreturn temperatures: 7°C and13°C
Approach temperature at thepre-condenser and rst-stagecondenser: 1.67°CDischarge pressure: 815 torr
Water vapour in the load gasis increased by 365% for Cases2C and 2D, compared to Cases2A and 2B. No pre-condenser(using chilled water) is used inCases 2A and 2C, whereaspre-condenser, cooled withchilled water, is used in Cases2B and 2D. For all the cases,inter-stage and after-condens-ers use cooling water. From
Table 2, it can be seen that theuse of chilled water in thepre-condenser reduced thesteam consumption by 28.5%in Case 2B compared to Case2A, and by 60.3% in Case 2Dcompared to Case 2C. It alsoreduced the total heat transferarea for pre-, inter- andafter-condensers by 30.8% inCase 2B compared to Case 2A,
and by 69.0% in Case 2Dcompared to Case 2C. Moresteam and system capital costreduction can be achieved by
using chilled water, if morewater vapour is present in theload gas.
Case study 3Case study 3 is the vacuumsystem optimisation of avacuum distillation column inan existing Asian renery.Vacuum required at the top ofthe column is 35 torr, the
vacuum system suction owrate is 10 070 kg/hr and thedischarge pressure is 895 torr.Suction gases contain 5641 kg/hr of water vapour, 3730 kg/hrof HC vapour (molecularweight = 172) and 699 kg/hr ofnon-condensable gases (molec-ular weight = 29). Details of theexisting vacuum system withfour steam ejector stages are
there are two ejectors in paral-lel in each of stages 2, 3 and 4.
The following alternativesare considered for retrottingto minimise the operating costsof the existing vacuum systemin the base case (Case 3A).
Cases 3B and 3CFour-stage steam ejector systemwith one pre-condenser (at the
inlet of the rst-stage steamejector), cooled by chilledwater. A single-stage absorp-tion chiller is used in Case 3Bfor generation of chilled waterusing very low-pressure ashsteam at 1.5 barg and 130°C. Amechanical chiller, using R143arefrigerant, is used for thegeneration of chilled water inCase 3C. COP of the chiller is
5.68. Details of Cases 3B and
3C are shown in Figure 4.
Case 3DThe rst two stages are steamejectors, and the last two stagesare replaced by one LRVP. Themodied vacuum system isshown in Figure 5.
Cases 3E and 3FFor a pre-condenser using
chilled water generated by asingle-stage absorption chiller(Case 3E) and a mechanicalchiller (Case 3F), the rst twostages are steam ejectors andthe last two stages are replacedwith one LRVP. This improvedsystem is shown in Figure 6.
Equipment costsCapital costs of steam ejectors
are derived from an available
reference chart.4 Costs forcondensers are estimated usingthe Capcost program (based onan Excel spreadsheet).10 Fixedcosts for absorption, mechanicalchillers and LRVP are based onvendor quotations. The installedcost for a mechanical chiller andLRVP were taken as twice thepurchase cost, whereas theinstalled cost for an absorption
(single-stage-LiBr) chiller wastaken as 1.5 times the purchasecost since the absorption chilleris a packaged unit and involvesless expensive installation.Costs for foul water treatmentare used from reference 5.Utility costs assumed are: MPsteam = $32.14/ton; LLP steam= $31.65/ton; cooling water =$0.05/ton; and electric power =
coolers and product coolers,which can also be used togenerate low-pressure steam. IfMP steam cost is lower (~75%compared to the base case, dueto credit from power genera-tion at steam turbines), amechanical chiller will not beeconomical. For this case, asingle-stage absorption chiller(Case 3B) will be feasible(payback ~3.3 years) only if thelow-pressure steam is availableat zero cost (see Table 3). Theoption involving a pre-condenser (cooled by chilledwater), rst two stages with a
steam ejector and the last twostages replaced by one LRVP(Cases 3F and 3E), is a veryattractive investment with apayback period of 1.8 to threeyears (with an incrementalcapital cost of $1.75 million to$3.25 million). Thus, replace-ment of the last two stages of amulti-stage steam ejectorsystem with one LRVP is
highly benecial.
ConclusionsThis article analysed the tech-niques available to reduce theenergy consumption of vacuumsystems used in petroleumreneries. Key requirements,
benets and constraints forimplementation of these tech-niques are highlighted, and
availability of waste low-pres-sure steam or low electricityprices.
Accurate estimation of avacuum system’s suction loadfor various plant operatingscenarios is difcult. Hence,considerable safety margin withrespect to suction load is oftenallowed in the system design.In this case study, the rst-stagesteam ejector is a single unit,hence it is very difcult toreduce the steam consumptionif the operating suction gas loadis lower than the design rate.However, for the case of a
steam ejector system with achilled water pre-condenser,electrical power required for thechiller can be reduced by capac-ity control or by operating afew chillers arranged in paralleloperation. Thus, operating cost
benets can be furtherincreased. Availability of plotspace and maintenance costsare the other critical issues for
installing a chiller. Installationof a mechanical chiller andLRVP may require modicationcosts at a power intakesubstation.
Waste LLP steam can berecovered economically fromthe steam condensate system.Petroleum reneries oftendiscard a lot of waste heatthrough furnace stacks, n-fan
are presented in Table 5. Forthe cases where operating costsincreased compared to the basecase, payback is shown as “noteconomical” in this table.
Analysis of the resultsAnnual operating costs of avacuum system are very signif-icant compared to the installedcosts. Hence, optimum vacuumsystem conguration is essen-tial to minimise operatingcosts. Even for the optimumdesign, the majority of theoperating cost arises from therst stage, which handles the
maximum ow rate of gases/vapours. The only way toreduce this cost is to condensethe vapours before they reachthe rst-stage ejector.Depending on the suction pres-sure of the rst stage, chilledwater may be required. For therenery case study, themechanical chiller option (Case3C) has a payback period of
5.28 years, which will reduce to2.25 years if the power cost islower at 75% compared to the
base case (see Table 3). If wastelow-pressure steam (~1.5 barg)is available at zero cost, asingle-stage absorption chiller(Case 3B) will have a paybackperiod of 2.45 years. Thus, the
benets of pre-condensersusing chilled water depends on
Values from optimisation
Utilities Pre-condenser Inter-condensers Total Number of Chilled Cooling Vent Vent Vent heat Inter-stage Pre- MP steam, water, water, pressure, Pre-condenser, temperature pressure, Inter- and after- transfer Case Stages condensers condenser kg/h kg/h kg/h mmHg area, m2 °C mmHg condenser area, m2 area, m2
2A 4 3 No 351 0 44 581 NA NA 33.9/41.7/48.9 76.5/157.6/280.9 31.4/4/2/1.4 3928 4 3 Yes 251 9011 25 346 36 11 33.9/41.7/48.1 110.6/203.4/329.6 10.97/2.4/1.5/1.2 27
2C 4 3 No 632 0 110 202 NA NA 33.9/41.7/48.1 55.9/119.1/226 150/6.9/2.4/1.35 16120 4 3 Yes 251 42 901 25 346 36 34.70 33.9/41.7/48.1 110.6/203.4/329.6 10.97/1.9/1.2/1 50
Effect of pre-condenser with chilled water cooling on steam consumption in the multi-stage steam ejector system(case study 2)
cic factors such as energy cost,plot size, capital cost, acceptablepayback period, operationalreliability, maintenance andsafety issues.
Nomenclature
Ai Area for the ith stage condenser,
m2
Dpi Process gas-side pressure drop
across the condenser at the outlet of the
ith ejector, torr
ation, thus improving energy
efciency and also reducingcarbon emissions. The econom-ics of such an optimisation varyfrom one site to another as thecosts of steam, power andextent of waste heat recoveryvary greatly. A detailedeconomic study similar to theone shown in the present studycan be conducted to decide the
best strategy for a particular
strategies for selection and
implementation of a suitablemethod are outlined. It can beconcluded from the analysisthat use of chilled water at thepre-condenser reduces theenergy costs of vacuumsystems. As reneries operatemany steam ejector vacuumsystems, considerable potentialexists for reducing energyconsumption for vacuum gener-
Details of the vacuum system Case 3A Case 3B Case 3C Case 3D Case 3E Case3FSteam ejectors MP steam, ton/hr 7.814 3.635 3.635 5.9 1.882 1.882Condenser area Total area, m2 2145 1893 1893 2020 1813 1813LRVP Power consumption, KW 0 0 0 174 174 174 Cooling water, ton/hr 0 0 0 30 30 30Single-stage Cooling water, ton/hr 0 1126 0 0 1126 0
absorption chiller LLP steam, ton/hr 0 10.58 0 0 10.58 0 Power consumption (includes 0 160 0 0 160 0 power for chilled water pumps), KWMechanical chiller Power consumption, KW 0 0 1010 0 0 1010 Cooling water, ton/hr 0 0 618 0 0 618Total utility requirements MP steam, ton/hr 7.814 3.635 3.635 5.9 1.882 1.882 LLP steam, ton/hr 0 10.58 0 0 10.58 0 Chilled water, ton/hr 0 642.8 642.8 0 642.8 642.8 Cooling water, ton/hr 1099 1197 689 1129 1227 719 Power requirement, KW 0 160 1010 174 334 1184Savings at foul water stripper LP steam saved at foul water stripper, ton/hr 0 0.72 0.72 0.33 1.02 1.02 Power saved at foul water stripper, KW/hr 0 11.41 11.41 5.23 16.19 16.19 Cooling water saved at foul water stripper, ton/hr 0 2.56 2.56 1.17 3.64 3.64Operating cost, million $/year Base case 2.68 4.48 2.44 2.29 4.13 2.10 Considering LLP steam for 2.68 1.54 2.44 2.29 1.20 2.10 absorption chiller is zero cost
Considering electric power cost is 2.68 4.43 2.11 2.23 4.03 1.71 75% of the cost considered If MP steam cost is 75% of that previously due 2.13 4.22 2.18 1.87 4.00 1.96 to credit from power generation at steam turbines If MP steam cost is 75% of that previously due 2.13 1.29 2.18 1.87 1.07 1.96 to credit from power generation at steam turbines and LLP steam to absorption chiller is at zero costCapital cost, million $ Steam ejectors 0.371 0.159 0.159 0.330 0.118 0.118 Surface condensers 1.691 1.691 1.691 1.497 1.497 1.497 LRVP 0.000 0.000 0.000 0.692 0.692 0.692 Single-stage absorption chiller 0.000 3.000 0.000 0.000 3.000 0.000 Mechanical chiller 0.000 0.000 1.500 0.000 0.000 1.500 Total installation cost 2.062 4.850 3.350 2.519 5.307 3.807Payback period Base case - Not 5.28 1.16 Not 2.98 economical economical If LLP steam for absorption chiller is - 2.45 5.28 1.16 2.19 2.98
available at zero cost If electricity cost is 75% of that - Not 2.25 1.01 Not 1.80 considered for the above cases economical economical If MP steam cost is 75% of that - Not Not 1.75 Not 10.37 considered in the base case (due to economical economical economical credit from power generation at steam turbines) If MP steam cost is 75% of that - 3.30 Not 1.75 3.05 10.37 considered in base case and LLP economical steam to absorption chiller is available at zero cost
Analysis of alternatives for the vacuum system of a vacuum distillation column in a petroleum refinery