It is clear that major resource andprocessing industries will soon haveto compete in an evermore

greenhouse gas (GHG) constrainedworld. While the exact timing and detailof some GHG legislation is stilluncertain, consultants generally advisethat proactive companies can “self-insure” and prosper by adoptinginnovative and cost-effective CO2-trimming technologies.

The processing industries havealready been operating for some timeunder restrictive legislation on NOX andSOX emissions. To provide a degree offlexibility on compliance, a market hasbeen set in place that trades emissionscredits for these two acid rain-formingchemicals. Those who have found waysto reduce their emissions below quotacan now profitably sell their excess“right to emit” to those who have not.This market is fairly active. For instance,in 2003 in the California and Texasmarkets there was consistent tradingfrom month-to-month in both theallowance and emission reductioncredits (ERC) markets. This trend shouldcontinue given the US EnvironmentalProtection Agency’s (EPA’s) mandate tofurther reduce emissions in bothHouston and Los Angeles. The two citieshave been competing for theunfortunate privilege of having theworst air quality of any majormetropolitan area in the US.

Heat recoveryExcellence in process heat recovery is animportant aspect in an oil refiner’soverall strategy for dealing withatmospheric emissions. In fact,replacing generated heat with recoveredheat should be a priority, because it isone of the few compliance toolsavailable with a positive return oninvestment (ROI).

In the reactor section of manyhydrotreating and reformer projects, thePackinox welded plate heat exchanger(WPHE) technology has significantlylowered both the operating expense of

fuel and the total life-cycle cost ofinstalled plants, relative to that possiblewith alternative exchanger types.Figures 1 and 2 illustrate WPHEs inreactor feed/effluent (F/E) service of low-pressure reformer and high-pressureHDS units.

WPHEs are built as a one-pass, truecounterflow plate pack inside a pressurevessel. The use of long plates allows for

tight thermal pinches between processstreams and permits optimised heatrecovery. The compactness of the designallows for large heat-transfer surfaceareas in relatively small exchangers.This keeps pressure drops low despitethe high heat recovery. WPHEs havebeen built with areas over 15 000m2

(160 000ft2) and duties up to about150MW (500MM BTU/hr). A WPHE

Saving fuel costs withWPHEs

WPHEs save fuel costs and serve as a cost-effective CO2-trimming technology byreducing heater duty in reformers and HDS units. The resulting fuel cost

savings add to emission trade benefits

Peter Barnes PBA Consulting Pierre Xavier Bussonnet Packinox SA François Reverdy Packinox Inc

REFINING

PTQ Q2 2005w w w. e p t q . c o m

1

Figure 1 Reformer reactor feed/effluent HE

09 packinox(FINAL).qxd 24/5/05 05:06 PM Page 1

operates comfortably up to 550°C(1000°F). Maximum operating pressureis determined only by the design of thesurrounding pressure vessel. WPHEshave so far been put in service at up to120 bars (1750psi).

Typical payout times are less than twoyears, sometimes less than one.Traditional fuel savings still representthe greatest added value of a WPHE.However, in the new clean air era, thespin-off benefits from CO2, NOX andSOX credits may add extra value,amounting to about 15–25% of the fuelcost savings of an energy-saving WPHEproject.

Kyoto overviewIn 1990, a group of eminent scientistsand Noble Prize winners reported that:“the growing greenhouse effect has thepotential to produce dramatic changesin climate.”1 The GHG emissions listincludes CO2, methane and some CFCs

(chloro-fluoro carbons). Excluded arewell-known acid rain or toxic airpollutants such as SOx, NOx andparticulates. For most practical purposesin discussions of GHG legislation,trading and project planning, GHGsimply means CO2.

The Rio Convention of 1992, signedby the US and 155 other nations, calledfor a non-binding goal to stabilise GHGemissions at 1990 levels by 2000. It waslater realised that most “Rio-signatories”could not or would not meet their 2000GHG targets, so in 1995 a Conference ofthe Parties (COP-1) called for a new,stronger but more doable deal to benegotiated. This led to the KyotoProtocol of 1997 (Kyoto). This treaty wassigned by more than 160 nations. Afterratification back home, a nation becomesa member state and is bound to meet itsagreed GHG targets. For each first-worldindustrialised country, the broad goal isto cut GHG emissions to about 7% below

its 1990 levels by 2012 (for example, 6%,7% and 8% below for Japan, the US andEU, respectively). Aiming for efficientprogress towards global GHG targets atthe lowest overall cost, Kyoto allowsindustry to harness global market forcesvia three flexible mechanisms, ineffect all variations of domestic orinternational emissions trading (ET).

Slowing but not stopping Kyoto’sprogress are some grey areas. Theseinclude debates on how to fairly allocatethe allowed global total CO2 emissionbetween first- and third-world countries;how to monitor GHG emissions (athome and abroad); how to manageinternational ET; and how to setcompany penalties for non-compliance.Expected deterrents are fines, plantclosures and even prison terms forsenior company officers. In 2003,certain US congressional supporters ofKyoto sponsored a bi-partisan billrequiring the US to reduce its GHGemissions to 2000 levels by 2010, andthen to 1990 levels by 2016. Meanwhile,several US companies are already cuttingGHG. For example, by improvingpollution controls and energyconservation, DuPont has slashed itsemissions to 60% below 1990 levels.

Despite all such impediments, workhas continued widely to resolvelingering disputes and to finishtranslating Kyoto’s broad ideals into thefine details of workable legislation. ByOctober 2004, 124 countries, includingthe EU, Russia, Japan, Canada,Switzerland and Norway, had ratifiedKyoto to make it supported by countriesthat emit at least 55% of the world’sGHG. At this level, it becameinternational law on 16 February 2004and will legally bind 55 signed-upindustrialised nations to makesignificant GHG cuts by 2012. Manyobservers believe that Koyto’s slow butsteady progress toward universalacceptance is now unstoppable. Theyhope and trust that non-member largeemitters like China and the US willeventually feel obligated to do their partin retarding global warming.

Kyoto: working detailsA simplified summary of Kyoto’semerging workplan is now offered. Theglue that holds it all together is the CO2

emission allowance (EA), also called aCO2 permit or credit. One EA confers theright to emit one metric ton of CO2 (orCO2 equivalent). In January 2008, at thestart of the five-year commitmentperiod (2008–2012), a signed-upmember state or government (G) will bedeemed to hold a stockpile of EAs basedon the 1990 CO2 levels it negotiated atKyoto in 1997.

A nation’s current accountable GHG

PTQ Q2 2005

2

REFINING

Figure 2 HDS unit reactor feed/effluent HE

09 packinox(FINAL).qxd 24/5/05 05:06 PM Page 2

PTQ Q2 2005

3

REFINING

level will be taken as the total annualCO2 emissions reported by all its largefinal emitters on a manageable list ofincluded industries in three mainsectors: oil and gas, electricity andmining/manufacturing. The hugetransportation sector is notionallycovered by regarding it as part of the oilindustry, whose fuel sales are deemed tobe fuel burned by the oil companies.Matching CO2 emissions are thenincluded in the oil industry’s annualCO2 tally. In this way, each developedcountry can adequately account forabout 75% of its total CO2 emissions.

From its EA stockpile, a G nowallocates or freely grants an appropriatenumber of its EAs to each includedcompany on its three-sector list. For the2008–2012 commitment period, theinitial EA numbers allocated in, say,December 2007 will be related to eachcompany’s 1990 CO2 emissions. For oiland gas companies, this includes all theCO2 emitted when their customersburned all refined fuel volumes sold bythe oil companies in 1990. For theselected large industries, CO2 emissionperformance is to be annually reviewedand improved. Such industries may buyor sell EAs by free market ET. Soon afterthe end of each commitment year(probably beginning around March–April 2009):— Each company must complete andreport to its G its CO2 emission recordfor the previous year. It must thensurrender to the G enough EAs to matchthe CO2 it emitted last year. In Europe,the proposed shortfall fine is €40 per EAdue but unreturned.3 A similar fine maywell apply in the US if/when Kyoto isratified or the M&L bill passes. As itapplies only to the CO2 margin abovepermit, the EU fine for EA shortfallsshould not be likened to the dreadedcarbon tax idea that would apply toevery ton of CO2 emitted— The G must delete or write off about1% of its EAs each year. This ratchetsdown the member state’s total CO2 EA,keeping the G on track to meet itsDecember 2012 Kyoto target. For eachsuccessive year, the G reallocates anever-diminishing EA stockpile to theincluded companies on its three-sectorlist after any required updates.

The ET programme for CO2 would besimilar to existing programmes for SOx

and NOx trading. Companies that trimtheir emissions below allowance levelcould sell their surplus EAs to firms thatexceed their emission limits. Thus,excessive emitters can buy cheap, tradedEAs to avoid costly shortfall fines, whilemore entrepreneurial industries have abonus incentive (beyond lower fuelcosts) to implement real energy-savingprojects.

The Europeans are currentlydeveloping a central electronic EAregistry system. This will hold all detailsof both national and company EAaccounts and will record details of EAownership changes due to emissiontrades. A balance will be sought betweenenvironmental transparency andcommercial confidentiality.

Free market forces alone willdetermine the price of traded EAs. Howclosely the traded EA price will approachthe shortfall fine (for example, $40) willdepend on EA supply and demand. Inturn, these two factors will depend overtime on:— The likelihood of Kyoto lawapplying in the location concerned — How well most local companies aremeeting their CO2 quotas.

While CO2 credits were traded in theUS during October 2004 at about $1 percredit, some US emission traders haveforecast market prices of around $20before 2008 (ie, about half the €40shortfall fine previously quoted). Thissuggests that many US oil industryanalysts foresee a 50:50 prospect of USlaws on CO2 emission within three orfour years, with a significant number ofincluded companies still tardy in theirenergy conservation efforts and thereforeobliged to bid for additional EAs. Bycontrast, companies proactive in energyconservation are foreseen to have anattractive market for their surplus EAs.

CO2 trade: the 15% ruleAs a general rule of thumb, CO2 creditsmay add extra value, amounting toabout 15% of the fuel cost savings of anenergy efficiency project. To quantifythe $ value of GHG reduction, handyguidelines have been developed asfollows:— For international readers and forbrevity, heat duties are stated inmegawatt (MW)

Conversion factors: 1MW = 3.413MMBtu/hour = 20 640tcal/day — To relate easily with annual GHGmass-flow accounting, both CO2 andhydrocarbon fuel flows are stated inmetric tonnes per annum (tpa) ratherthan in imperial volumes per hour.

For natural gas (NG) fuel, broad-brushestimates may be derived per MWreduction in charge heater radiant cellduty via data and equations as follows:

Flows of NG and CO2 saved:LHV, natural gas = 11 500tcal/tOn-stream factor = 350 days paNG/MW released = 20 600/11500*350

= 627tpa/MWAt radiant cell eff’y = 65% NG fuel saved = 627/0.65

= 965tpa/MW (1)

Carbon/typical NG = 71 %wt CO2/NG burned = 44/12 * 71%

= 2.6 ton/tonCO2/radiant duty = 965 * 2.6

= 2500tpa/MW (2)

Gas price G $/t related to crude $/bbl:LHV of crude oil = 10 400tcal/tLHV ratio NG/crude = 1.106Crude oil sp Vol = 7.2bbl/tonCrude price/barrel = B $/bblAt same $/btu:NG price/ton: G = B* 7.2 * 1.106

G = 8 * B $/ton(3)

Annual $ benefits per MW:NG cost saved = 965 * 8 * B

= B * 7700 $pa (4)

CO2 trade price = Ec $/ton CO2 trade value = Ec * 2500 $pa

(5)

Typical example values/MW:Say crude oil B = 40 $/bbl. And:CO2 credits Ec = 20$/t. Then byEq 4.5:Fuel cost saved = 308 000 $pa/MW

(6)CO2 trade value = 50 000 $pa/MW

(7)

Thus, when Packinox cuts chargeheater radiant cell duty by “Q” MW,typical fuel cost savings are about $Q *300 000 pa. However, for Kyoto memberstates, CO2 credits are a significantbonus, adding roughly another $Q * 50 000 pa of profit.

SOx and NOx trade: the 10%ruleAs a rough rule of thumb, combined SOx

and NOx credits (when applicable) canadd value amounting to around 10% offuel cost savings for an energy efficiencyproject, as derived in the followingdiscussion. (Unlike the 15% for CO2

credits, this 10% estimate may varywidely — for example, 0–20% —depending on the sulphur and nitrogencontents of the fuel burned.)

Molecules of SOx & NOx cause localacid rain rather than global greenhousewarming. Thus, for SOx andNOx, emission legislation generallyfocuses on some perimeter-definedbubble, not on the whole planet.However, as such differences do notimpede the SOx or NOx ET, refiners thatuse Packinox equipment to cut fuel billsmay also benefit from such trade. Ifemitting below allowed SOx caps, therefiner can sell surplus SOx permits. Ifemitting above caps, decreasingemission will save the refiner thepurchase cost of SOx permits. Either way

09 packinox(FINAL).qxd 24/5/05 05:06 PM Page 3

gives about the same ET benefits.For quick scouting studies on SOx and

NOx trade incentives to reduce chargeheater duty, simple steps (at constantfuel price per heat unit) may beemployed as follows:

All sulphur (atomic wt = 32) in fuelburns to SOx, calculated as SO3 (mol.wt= 80). Fuel gas is typically amine treatedand sweet fuel oil (FO) is usually themain source of oil refinery SOxemission. For a fair scouting estimate (inthe absence of any flue gas sulphur-removal system), factors per MWreduction in charge heater radiant cellduty may be derived as follows:

Flows of oil and SOx saved:LHV, fuel oil = 10 000tcal/t FO/MW released = 20 600/10 000 *350

= 720tpa/MWAt radiant cell eff’y = 65%:Fuel oil saved = 720/0.65

= 1100tpa/MW (8)

Sulphur/fuel = S %wtSOx/S ratio = 2.5t/t (= 80/32)SOx reduction = 1100 * S/100 * 2.5

= 27*S tpa/MW (9)

Economic incentive: SOx ET price = Es $/tonSOx ET value = 27*S*Es $pa (10)

Typical example values/MW:For SOx ET, Es = 1000$/t& FO sulphur, S = 0.6 %wt*

SOx trade value = 16 200 $pa/MW(11)

Total NOx has two main sources :— Thermal NOx: high temperatureoxidation of molecular N2 in pure air— Fuel NOx: direct oxidation of organicnitrogen (for example, NH3) in fuel.

These are separately calculated, thenadded to give total NOx make. Finally,this sum is corrected for any post-combustion NOx control technology,such as selective catalytic reduction ofacidic NOx to benign N2, to give actualNOx emission to atmosphere.

Thermal NOx is primarily a functionof flame temperature (Tf). In turn, Tf is acomplex function of fuel LHV,stoichiometric air/fuel ratio, excess air,air preheat, flue-gas recirculation,humidity, low-NOx burners, burnerintensity, heat-removal rate and gas-mixresidence time. Often, in practice, somefactors are trivial or absent and may beneglected.

ExxonMobil R&D reports acomprehensive calculation method forthermal NOx.

4 This takes into accountall the previously noted factors, but isoutside the scope of this discussion.However, for quick ET scouting studies,Equation 9 gives a rough estimate forthermal NOx emission in tpa per MWradiant cell duty. This simulates typicalmodern heaters on-line for some350sd/yr with about 150°C air preheat.

Thermal NOx = 3 +/-1.3tpa/MW (12)

The user may choose to “gestimate”small variations on either side of thebase number if they know of somefactor that will increase or decreasetypical Tf (for example, large hydrogencontents of fuel gas will raise Tf andthermal NOx, and absence of air preheatwill lower Tf and thermal NOx).

The fuel NOx calculation assumes thatall N (at.wt = 14) burns to NO2 (mol.wt= 46). Then, approximate factors perMW reduction in charge heater radiantcell duty may be derived as follows:

Fuel/radiant duty = 1100tpa/MW Nitrogen/fuel = N %wt.NOX/N ratio = 3.3t/t (= 46/14)NOX reduction = 1100 * N/100 * 3.3

= N * 36tpa/MW (13)

Total NOx economic incentive:NOX ET price = En $/tNOX ET value = (3+36*N)* En $pa/MW

(14)Typical example values/MW:For fuel nitrogen N = 0.25 % andNOx trade price, En = 1000$/t:

NOx value by Eq 14 = 12 000 $pa/MW (15)

Combined benefit of SOx andNOx creditsPer MW reduction in radiant cell duty,the previously noted example values

PTQ Q2 2005

4

600

500

400

300

200

100

0

0 20 40 60 80 100 120

Heat release

curves, MW

Temp

deg. C

Effluent

Dew pt. 137 ºC

Pinch = 35 ºC

Feed mix

Dry pt. 156 ºC

Liq. fd t/d/10

Liq. eff t/d/10

HAT = 48 ºC

CAT = 38 ºC

Exch'r = 81.8 mw

Heater = 12.1 mw

Cooler = 11.6 mw

Reformer type :

F/E heat exch'r :

Feedrate = 3000 t/d

Rx-outlet = 12 barg

Rec.gas = 1100 t/d

Fixed bed Shell & tube

Figure 3 F-bed reformer, S&T

REFINING

09 packinox(FINAL).qxd 24/5/05 05:06 PM Page 4

would provide total SOx and NOx creditsof almost 30 000 $pa/MW. As before,fuel cost savings would be some 300 000$pa/MW. Thus, the notional ratio ofpotential SOx and NOx credits/fuelsavings is about 10%.

How a WPHE saves fuelIn refinery oil conversion orhydrotreating units, WPHEs are widelyemployed as reactor F/E heat exchangersin the high-pressure (HP) reactor loop.In such loops, the fired heater’s key roleis to inject a slug of high-level heat intoreactor charge, essentially via theradiant cell. However, radiant heattransfer is typically only around 60% ofthe total heat released by fuelcombustion. The remaining 40% iswaste heat, of which about half may berecovered at a relatively low temperaturelevel (for example, in a waste heatboiler, WHB), while the rest is directlywasted as hot stack gas to theatmosphere.

In most refineries, abundant low-levelheat sources (such as in hot flue gas, hotcolumn top-vapours and uncooledproduct rundowns) greatly outweigh thescarce low-level heat sinks (such asheavy oil tanks and lines). When WHBsteam from avoidable hot flue gasultimately condenses against a low-levelheat sink, the warming of that sinkusually dooms the potential recovery oftruly unavoidable waste heat fromelsewhere. Thus, in the final analysis,marginal heat recovery in a plant WHB

may have little impact on total refineryfuel consumption. Consequently,within the accuracy of carbon creditarithmetic, the long-term fuel burn andCO2 result of changes in radiant cellduty (or reactor charge heat duty) maybe computed as if the fired heater hadno WHB. Thus, we may take fired heaterefficiency = radiant cell efficiency =60%, or 65% for a more conservativeanalysis.

The following case studies show howcharge heater duties are reduced byachieving much closer pinchtemperatures (or delta-Ts) within the F/Eplate exchanger than practicallypossible with tubular exchanger types.As shown in the example heat-releasecurves of Figures 3 to 6, this also meanssmaller HAT and CAT values (hot-endand cold-end approach temperatures)for the HE. Thus, the charge heater’sinlet temperature is hotter, its heat dutyis lower, so less fuel is burned and lessCO2, SOx and NOx are emitted. Pushingthe concept to its logical extreme, oneparticular Packinox-based HDS-typeunit with a high reactor exotherm nowoperates well with its charge heatertotally bypassed and thus its stackemissions cut to zero. In all cases, thesmaller CAT value means the requiredcooler duty is also lower, with furthersavings in capital cost and plot space.

Low-pinch technologyThe vital small-pinch results from thecombination of:

— Carefully managed, uniformdistribution of two-phase flow: F/Einlet lines have patented liquid/vapourpre-mixing devices. The wide-slot entryto all flow canals plus static mixer typeflow homogenisation within the platepack ensures well-mixed turbulent flowat every point in the plate pack, with notroublesome dead zones— High overall heat-transfer co-efficient (OHTC): Packinox OHTC isabout twice that of S&T exchangers dueto the above flow pattern pluscorrugations on Packinox plates that areoptimised to promote useful turbulencewith minimal pressure drop — Very large heat-transfer surfaceareas: compact plate pack constructionfacilitates more m2 area per m3 of shellvolume.

In numerous cases, the overall effectof these three factors has allowedrefiners to get a larger guaranteed heatduty (and a lower pressure drop) from asingle HE than would ever have beenpossible from multiple S&T exchangers.The Packinox option also savesinstallation costs by simplifying pipingand using less plot space.

Case 1Fixed-bed reformer: F/E exchanger

upgradeIn one early 1990s revamp for capacityincrease, an Asian reformer’s F/E HEsystem was upgraded by replacing itsearly model Texas Tower with a moreefficient WPHE. This avoided an

0

100

200

300

400

500

600

0 20 40 60 80 100 120

Heat release

curves, MW

Temp

deg.C

Effluent

Dew pt. 137 o

C

Pinch = 10 o

C

Feed mix

Dry pt. 156 o

C

Liq. fd t/d /10

Liq. eff t/d /10

HAT = 12 o

C

CAT = 11 o

C

Exch'r = 88.7 mw

Heater = 5.1 mw

Cooler = 4.6 mw

Reformer type :

F/E heat exch'r :

Feedrate = 3000 t/d

Rx-outlet = 12 barg

Rec.gas = 1100 t/d

Fixed bedPackinox

Figure 4 F-bed reformer, PHE

REFINING

PTQ Q2 2005

5

09 packinox(FINAL).qxd 24/5/05 05:06 PM Page 5

expensive charge heater project plus adifficult and costly extension of theproduct cooler with limited availableplot space. Total capital cost of thePackinox option was significantly lessthan the heater-plus-cooler-basedalternative revamp.

Figures 3 and 4 show key reactor loopconditions plus F/E exchanger HRCsbefore and after the revamp. For about ayear after the revamp, the unit was onlyrequired to run at its original 3000t/dcapacity. Due to F/E pinch pointimproving from 35–10°C, charge heaterduty decreased from 12.1–5.1MW,lowering radiant cell duty by 7MW. Thecorresponding drop in fuel gasconsumption was about 6700tpa (seeEquation 1).

At cost parity with 40 $/bbl crude, thefuel cost savings amount to $2.1 millionpa (= 7MW * 308 000 $pa/MW, Equation6). This paid out the WPHE in a shorttime and continues the refiner’s annualopex benefit to this day.

The post-revamp fuel reduction perMW was actually greater than theprevious discussion predicted. This isbecause, with the charge heater radiantcell now at below half its designcapacity, flue gas outlet temperaturefrom the radiant cell fell enough to raiseeffective heater efficiency from 60–70%.In the decade since this Case 1 revamp,F/E exchangers have been furtherimproved so that now Texas Towers

claim pinch points of 22–25°C, whileWPHEs compete today with pinchpoints of 7–10°C. Today, delta emissioncredits may be predicted by Equations 7,11 and 15 as follows (Note: thefollowing example calculation assumesthat refinery fuel price equates to crudeat $40/bbl, S/fuel = 0.6%, N/fuel = 0.2%,and applicable CO2/SOx/NOx credits are20/1000/1000$/t):CO2: 7MW * 50 000 $pa = $350 000 pa SOx: 7MW * 16 000 $pa = $112 000 pa NOx: 7MW * 12 000 $pa = $84 000 paTotal emission credits = $546 000 pa

For input data other than assumed inthis discussion, these estimates can beeasily reworked using the simpleequations provided.

Case 2 CCR reformer: F/E exchanger upgrade

To compare reformer types, this studyexamines a revamp similar to Case 1,but with a more modern CCR reformerwhose F/E exchanger HRCs are shown inFigures 5 and 6. It is assumed that theCCR upgrade options will be either amodern Texas Tower with a 22°C pinchpoint or a Packinox HE with a 7°C pinchpoint.

At the same naphtha charge rate of3000t/d, this unit runs at about half theROP (6 not 12 barg) and a much lowermol-ratio, since recycle gas (RG) flowrate is much lower (250 not 1100t/d).

This much lower vapour-to-liquid

ratio, coupled with the requirement forlower pressure drop, has made it a lotmore difficult to consistently achieveproper dual-phase feed distributionand uniform, vigorous liquid lift insidethe heat-transfer bundle. In thatrespect, the static mixer effect generatedby the geometry of plate heatexchangers has become a significantasset in the drive toward lowerhydrogen-to-oil ratios.

The Packinox option now reducesheater duty by an additional 3.4MW(9.2–5.8), with fuel savings of about$1.05 million pa (= 308 000 * 3.4, byEquation 6). Ongoing million $pasavings such as these clearly make aWPHE (rather than a S&T) the bettereconomic option for any CCR upgrade.

Using the same method as in Case 1,emission credits are predicted as follows:CO2: 3.4MW * 50 000 $pa = $170000 paSOx: 3.4MW * 16 000 $pa = $54 000 pa NOx: 3.4MW * 12 000 $pa = $41 000 paTotal emission credits = $265 000 pa

Case 3Gas oil HDS redesign

In the mid-1990s, a large number ofHDS studies were commissioned forinterested refiners in Asia, Europe andthe Americas. For the average studycapacity of 35 000bpd, the five-yearbenefit was $10 million for selectingPackinox PHEs rather than S&Texchangers. This equates to just under

00

100100

200200

300300

400400

500500

600600

0 20 40 60 80

Heat release

curves, MW

Temp

deg.C

Effluent

Dew pt. 146 oC

Pinch = 22 oC

Feed mix

Dry pt. 186 oC

Liq. fd t/d /10

Liq. eff t/d /10

HAT = 55 oC

CAT = 43 oC

Exch'r = 57.7 mw

Heater = 9.2 mw

Cooler = 9.6 mw

Reformer type : Cts. Cat Regen

F/E heat exch'r : Shell & Tube

Feedrate = 3000 t/d

Rx-outlet = 6 barg

Rec.gas = 250 t/d

Figure 5 CCR reformer, S&T

REFINING

PTQ Q2 2005

6

09 packinox(FINAL).qxd 24/5/05 05:06 PM Page 6

$3 million per 10 000bpd for a mix ofhot and cold HP separator designs.

In 1996, a major Taiwanese refiner(FPC) was planning two new 65 000bpd(= 8600t/d) cold separator HDS units.For all reactor and stripper F/E duties,FPC selected PHEs (rather than S&Texchangers). After a careful redesign byIFP, it was confirmed this would save$48 million in the refiner’s five-yearbusiness plan. In effect, this 1996 designstudy compared IFP alternatives (seeFigures 3 and 4, PTQ Spring 2004, p87, withthe referenced Figure 4 heater-duty correctedto 5MW) as follows.

For each plant, a 20MW energysaving (direct fuel and electricity) wasvalued at $3.3 million pa, or $16 millionfor the first five years of operation. Inaddition, fewer and smaller equipmentitems would cut installed capital cost by$8 million. Thus, total five-year savings= (16 + 8) x 2 = $48 million for the twounits. This equates to $3.7 million per10 000bpd (ie, close to the previouslymentioned study average).

Compared with the 23MW heater inthe superseded S&T option of Figure 3(PTQ Spring 2004), FPC’s project nowoperates successfully as per the Figure 4from PTQ Spring 2004, with chargeheater duty reduced by 18MW to only

5MW (Des, 12MW). Recycle gascompressor power is reduced by1.5MWe, equivalent to about 2MWradiant cell duty. In the foreseen Kyotoera, the net reduction of 20MW inradiant cell duty would be now creditedwith saving future fuel costs of about$6.2 million pa (= 20 * $308 000).

Where applicable, CO2 emissionpermits would be worth about $1million pa (= 20 * 50 000). Then, SOx

plus NOx permits would add possiblyanother $0.6 million pa. (This emergesby combining Equations 11 and 15 toobtain 20MW x $30 000/MW). Asestimated earlier, scouting-type totalemission credits of $1.6 million pa areroughly 25% of the $6.2 million pa totalfuel cost savings.

Of the $8 million capex saved perHDS unit at FPC, only $1 million wasdirectly due to lower F/E exchangerpurchase costs compared with S&Tcosts. The other $7 million was due tospin-off effects, such as smaller firedheaters, coolers and compressors, pluslower civil and piping installation costs.This project highlights a common issuefound on many Packinox projects,where the price difference between PHEand S&T equipment is relatively trivialbut hides much larger savings

elsewhere.

The authors extend their appreciation tothe editorial and technical contributionsfrom Daniel Sloan, president, EmissionReduction Specialists in the USA.

References1 Greenhouse gas trading — brief history,

Cantor Environmental Brokerage.2 Greenhouse gas emission curb, Chemical &

Engineering News, 13 January 2003. 3 Q&A on emissions trading and national

allocation plans, Europa — Rapid — PressReleases, Brussels 4 March 2004.

4 Takacs T J, Juedes D L, Crane I D, Methodestimates NOx from combustion equipment,Oil & Gas Journal, 21 June 2004.

Peter Barnes is chief process engineer, PBAConsulting in Australia, which he joined in1994 after retiring from Shell, where hiswork had included development andtechnical service for FCC andnaphtha/light ends upgrading.Pierre Xavier Bussonnet is technicaldirector for engineering and R&D withPackinox SA in France. François Reverdy is vice presidentmarketing, Packinox Inc in the USA. Hereceived his civil/mechanical engineeringdegree in 1979 in Europe and hisprofessional engineer’s licence in 1986 inthe state of Maryland, USA.Email: freverdy@packinoxinc.com

0

100

200

300

400

500

600

0 20 40 60 80

Temp

deg.C

Heat release

curves, MW

Effluent

Dew pt. 146 oC

Pinch = 7 oC

Feed mix

Dry pt. 186 oC

Liq. fd t/d /10

Liq. eff t/d /10

HAT = 29 oC

CAT = 27 oC

Exch'r = 61.1 mw

Heater = 5.8 mw

Cooler = 6.2 mw

Feedrate = 3000 t/d

Rx-Outlet = 6 barg

Rec.Gas = 250 t/d

Cts. cat regen Reformer type :

F/E heat exch'r : Packinox

Figure 6 CCR Reformer, PHE

REFINING

PTQ Q2 2005

7

On March 18, 2005 ALFA LAVAL has acquired PACKINOX which will change

its name to ALFA LAVAL PACKINOX.

09 packinox(FINAL).qxd 24/5/05 05:06 PM Page 7