Småskalig biokraftvärme från externeldad gasturbin - systemstudier, värmeväxling och turbinstyrning Sven Hermansson, Anders Hjörnhede, Daniel Ryde Per-Olof Sjögren Christoffer Boman, Jonathan Fagerström, Anders Rebbling Joseph Olwa, Marcus Öhman Per Nockhammar, Mattias Svensson SP Rapport 2015:64 SP Sveriges Tekniska Forskningsinstitut
70
Embed
Småskalig biokraftvärme från externeldad gasturbin ...ltu.diva-portal.org/smash/get/diva2:997683/FULLTEXT01.pdf · men är väsentligt lägre än både ORC, stirlingmotor och gasmotor.
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.
Transcript
Småskalig biokraftvärme från externeldad gasturbin - systemstudier, värmeväxling och turbinstyrning
Sven Hermansson, Anders Hjörnhede, Daniel Ryde Per-Olof Sjögren
Christoffer Boman, Jonathan Fagerström, Anders Rebbling Joseph Olwa, Marcus Öhman
Per Nockhammar, Mattias Svensson
SP Rapport 2015:64
SP
Sveriges T
eknis
ka F
ors
kn
ingsin
stitu
t
Småskalig biokraftvärme från extern-eldad gasturbin - systemstudier, värme-växling och turbinstyrning
Sven Hermansson, Anders Hjörnhede, Daniel Ryde, Per-Olof Sjögren, Christoffer Boman, Jonathan Fager-ström, Anders Rebbling, Joseph Olwa, Marcus Öhman, Per Nockhammar, Mattias Svensson
Abstract
Small scale CHP using externally fired hot air turbines - system
integration, heat exchanging and turbine control
Externally Fired Gas Turbines (EFGT) integrated with biomass combustion is a promising method for
small-scale combined heat and power production (CHP). However, for the technology to reach market
ready level, its potential need to be further established and research on system integration, high tem-
perature heat exchanging and turbine development need to be performed. In this project, the potential
of EFGT from 1-10 MWth Swedish woody biomass boilers, using an add-on concept with an under-
sized turbine, was assessed to 100-400 MWe. Furthermore was the use for electricity production from
waste gas of too low quality for gas engines assessed as very potent. The marginal electricity produc-
tion cost (the extra cost associated with the electricity production) would then be 1.1-1.3 SEK/kWh
(0.12-0.14 €/kWh), which is approximately 30-50% lower than other relevant technologies. To reach
the higher indicated levels of potential and the lower levels of production costs, thermodynamic simu-
lations further indicates that that care should be taken to the system integration of heat exchanging and
turbine operation. The exhaust turbine air should preferably be used as combustion air in the biomass
furnace, the turbine cycle should be humidified and the flue gas of the biomass boiler should be recir-
culated. An auxiliary alternative to increase the power output is to support fire the turbine cycle with
natural gas or biogas. This could also be used to modulate the operation of the system to balance for
load changes, which is believed to be of high importance in the future intermittent electricity system.
The experimental trials of the project show that depositions on the high temperature heat exchanger of
the EFGT system fired with stem wood pellets will be dominated by potassium sulfates, with traces of
chlorine. The same trials indicate that the release of the deposition substances potassium and sodium
from the fuel could be reduced with 75 and 95 % respectively, by combining a low combustion tem-
perature with additivation of the fuel. The exposure of materials to synthetic depositions and relevant
heat exchanging temperatures, indicated however that all selected alloys were more or less affected by
corrosion. However, a positive result was that pure potassium sulfate, which was found to dominate
the real depositions, caused little or no corrosion. Weighting these test results with material costs,
253MA was found to be a reasonable first selection. The heat exchanger design calculations then show
that 440 tubes of 2.5 m lengths would be needed for the biomass fired 100 kWe unit specified within
this project. The recommended position of the heat exchanger was found to be the upper part of the
shaft connecting the furnace with the boiler. Here, the temperature levels were found to be in the range
of the design temperature, however should the operational strategy be to maintain nominal load in or-
der to keep this temperature level. Finally were important steps made within the project to the integra-
tion of the gas turbine with a biomass furnace, by the development of a new control system for the tur-
bine cycle, including new software. By this action is the turbine system, with some further minor ad-
justment and completions, basically ready for application on a real process.
4 Systemstudier externeldad gasturbin 8 4.1 Objectives and method 8 4.2 Small-scale CHP cycles 9 4.3 The investigated EFTG system 10 4.4 Cost estimates 15 4.5 Thermodynamic model tests - methodology and simulations 19 4.6 Sensitivity analysis of component efficiencies 28 4.7 System power potentials 32
5 Högtemperaturvärmeväxling i rökgaser från biobränslepanna 38 5.1 Mål och genomförande 38 5.2 Experimentella studier av alkaliavgång och beläggningbildning 38 5.3 Experimentella studier av beläggnings- och materialinteraktioner 41 5.4 Designkriterier för värmeväxlare 48 5.5 Drift- och placeringsstrategi för värmeväxlare 51
6 Styrning av turbin samt integrering med befintlig panna 54 6.1 Mål och genomförande 54 6.2 Styrsystemets komponenter och dess uppbyggnad 54 6.3 Kontrollparametrar 55 6.4 Existerande styrsystem 55 6.5 Existerande kraftelektronik 56 6.6 Resultat 56 6.7 Slutsatser, måluppfyllelse och fortsatt arbete 59
7 Slutsatser 60
8 Referenser 64
3
Förord
Projektet har genomförts som ett samarbete mellan ett institut (SP Sveriges tekniska forskningsinsti-
tut), tre universitet (Luleå Tekniska Universitet, Umeå Universitet och Chalmers) och tre företag
(ENERTECH AB Osby Parca, Ecergy AB och MEGTEC Solutions AB). Projektledare har varit SP
Sveriges tekniska forskningsinstitut.
Projektet har finansierats av Energimyndigheten, inom ramen för Bränsleprogrammet Omvandling,
ENERTECH AB Osby Parca, Ecergy AB samt MEGTEC Systems AB.
4
Sammanfattning
Inom närvärmesektorn (< 5 MWth) finns en ökande efterfrågan på biokraftvärme. En av de ledande
tillverkarna, Enertech AB Osby Parca, får kontinuerligt förfrågningar från kunder som köper hetvat-
ten- och ånganläggningar om kraftvärmemöjlighet. Företaget har därför identifierat externeldad
gasturbin (Externally Fired Gas Turbine - EFGT) som en potentiell ny marknadsnisch för det småska-
liga segmentet (< 1 MWe). EFGT innebär att trycksatt luft värmeväxlas mot heta rökgaser inne i
eldstaden i en förbränningsanläggning, varefter luften expanderar genom en turbin kopplad till en ge-
nerator för elproduktion. I enlighet med Osby Parcas analys har tidigare studier identifierat EFGT som
ett ekonomiskt lovande alternativ för småskalig kraftvärme. Dock behövs potentialen för EFGT kart-
läggas mer ingående med avseende på systemintegration för maximal verkningsgrad, samt dess nisch i
det svenska energisystemet utvärderas i konkurrens med andra teknikval. Dessutom behövs teknisk
utveckling, främst inom högtemperaturvärmeväxlingen och turbinutvecklingen. I detta projekt har
därför ett brett konsortium med kompetens inom alla centrala delar i EFGT-systemet tagit sig an de
centrala frågor som rör småskalig kraftvärmeproduktion från biobränsledriven EFGT. Studierna har
genomförts på ett holistiskt sätt där både tekniska, systemmässiga och marknadsmässiga frågeställ-
ningar har undersökts samtidigt. Detta angreppssätt bedöms vara mycket viktigt för att möjliggöra pa-
rallell utveckling av kritiska tekniska komponenter, integrering av kraftcykeln mot värmekällan samt
att kunna identifiera var i energisystemet som tekniken är mest lämplig, utan att oönskade suboptime-
ringar åstadkoms.
Den övergripande genomgång som gjorts inom projektet av biobränsleförbränningsanläggningar i ef-
fektintervallet 1-10 MWth, vilka anses mest relevanta för småskalig biobränsleeldad EFGT, ger be-
dömningen att dessa skulle kunna bidra med ca 100-400 MWe, beroende på val av systemintegrations-
teknik. Kostnadsanalysen för investering i EFGT-cykel och drift visar dessutom på god effektivitet
jämfört med andra aktuella tekniker. Marginalkostnaden för investering i en EFGT-cykel för integre-
ring mot en biomassaanläggning bedöms uppgå till cirka 28-37 kSEK/kWe, beroende på tekniskut-
vecklingsnivå. Detta motsvarar ungefär samma kostnad som för marginalkostnaden för en ångcykel,
men är väsentligt lägre än både ORC, stirlingmotor och gasmotor. I jämförelsen av marginalkostnad
för elproduktion från respektive teknik står sig även här EFTG-tekniken väl. Här bedöms att marginal-
kostnaden för att producera 1 kWh el, förutsatt att behovet finns av att leverera resulterande mängd
värme, uppgår till ca 1,1 – 1,3 SEK/kWhe. Detta är uppskattningsvis 30-50 % lägre än motsvarande
marginalkostnad för ORC, gasmotor eller lågtrycksånga kombinerat med skruvexpander. Vidare visar
en modelleringsstudie av den tekniska systemintegrationen av en 100 kWe EFGT-cykel i 3000 kWth
biobränsleförbränningsanläggning att marginalelverkningsgraden (effektiviten i den extra bränslekon-
sumtion som krävs för att producera elen) kan fås att bli mycket hög, uppemot 70 %. Emellertid blir
det stora skillnader i effektivitet, beroende på hur tekniksystemet byggs upp och systemintegrationen
görs. Störst positiv inverkan visades återföring av turbinluften ha till biobränsleanläggningen som för-
bränningsluft, samt återföring av rökgaser från biobränsleförbränningen till förbränningsanläggningen.
Ett ytterligare intressant alternativ visade sig vara att stödelda turbinen direkt med ett gasformigt
bränsle, exempelvis biogas, efter värmeväxlingen mot biopannan. Då kan turbininloppstemperaturen
ökas över den temperatur som annars begränsas av värmeväxlingen, vilket medför att elverkningsgra-
den ökar. Denna påbyggnadsmöjlighet kan då användas för att modulera driften och leverera extra el
till nätet när det behövs som bäst, vilket är något som tros bli mycket viktigt i det framtida intermit-
tenta elnätet.
De delar av projektet som fokuserat på högtemperaturvärmeväxling och turbinsystemutveckling, visar
bland annat att beläggningarna på högtemperaturvärmeväxlaren vid förbränning av stamvedpellets
förväntas domineras av kaliumsulfater med en mindre mängd klor. Men labbförsöken visar även att
det finns stor potential att minska avgången av de beläggningsdrivande ämnena kalium och natrium
från bränslebädden med så mycket som 75 % respektive 95 %. Detta kan göras genom att använda
kombinationer av optimerade processparametrar (sänkt bäddtemperatur) och bränsleadditiv (kaolin el-
ler ammoniumsulfat). De materialexponeringar som syftade till att undersöka olika legeringars käns-
lighet för korrosion vid den temperatur som värmeväxlingen sker, uppvisade alla någon grad av hög-
5
temperaturkorrosion. Utifrån resultaten anses emellertid 253MA vara ett rimligt val av legering för en
första bedömning och fortsatta design och konstruktionsstudier, m a p kostnadsnivå kontra relativt god
korrosionsbeständighet. Emellertid förutsätts då att KCl inte riskerar beläggas på ytorna, då detta vi-
sade sig ha mycket negativ effekt på alla testade legeringar. En intressant observation i detta samman-
hang var emellertid att korrosionspåverkan för legeringar utsatta för ren K2SO4 var försumbar eller re-
lativt låg. Detta är att betrakta som positivt, med tanke på att beläggningsstudien visade att just K2SO4
dominerar beläggningarna. Slutligen konstateras utifrån givna kriterier för design, att värmeväxlaren
kan konstrueras som en tubvärmeväxlare med rökgaspassagen på insidan av tuberna. Med en tublängd
om 2,5 m krävs då 440 st tuber för att överföra den erforderliga energin till turbincykeln. Temperatur-
mätning i Osby Parcas pannteknik visar dessutom att värmeväxlaren lämpligen bör placeras i de övre
delarna av det vertikala schakt som sammanbinder slutförbränningskammaren och hetvattenpannan.
Den erforderliga temperaturen för värmeväxlingen visade sig uppnås här. Emellertid visar även mät-
ningen att lämplig driftstrategi för förbränningsanläggningen är att den körs på minst nominell drift för
att den erforderliga temperaturnivån ska kunna uppnås. Slutligen har inom projektet viktiga steg tagits
för att praktiskt möjliggöra integrering av hetluftsturbin tillsammans med biobränslepanna. För detta
har ett nytt styrsystem med ny mjukvara tagits fram för tubincykeln. Systemet bedöms med mindre
anpassningar och kompletteringar vara färdigt att applicera på processen.
6
1 Bakgrund
Inom närvärmesektorn finns en ökande efterfrågan på biokraftvärme. En av de ledande tillverkarna,
Enertech AB Osby Parca (EOP) anger att man kontinuerligt får förfrågningar från kunder som köper
hetvatten- och ånganläggningar om kraftvärmemöjlighet. Företaget har därför själv identifierat extern-
eldad gasturbin som en potentiell ny marknadsnisch för det småskaliga segmentet (< 1 MWe). Extern-
eldad gasturbin innebär, något förenklat och illustrerat i Figur 1, att trycksatt luft värmeväxlas mot
heta rökgaser inne i eldstaden i en förbränningsanläggning, varefter luften expanderar genom en turbin
kopplad till en generator för elproduktion.
Figur 1. Schematisk skiss över externeldad gasturbin i kombination med biomassaförbränning.
Värmeforskrapport 1237 [1] indikerar att EOP är inne på rätt väg. I rapporten bedöms att externeldad
gasturbin (Externally Fired Gas Turbin - EFGT) har lägst investeringskostnad och lägst specifik elpro-
duktionskostnad av aktuella teknikval (traditionell ångcykel, ångcykel med omvänd skruvkompressor,
ORC och gasmotor efter uppströmsförgasning) för småskalig kraftvärme från biomassaförbränning.
Alfavärde och elverkningsgrad bedöms dessutom kunna bli högre eller i samma storleksordning som
de jämförbara teknikerna (gasmotor). Det är dock uppenbart att potentialen för EFGT behövs kartläg-
gas systematiskt genom att undersöka i detalj hur systemet ska integreras för att uppnå maximal verk-
ningsgrad, samt dess nisch i det svenska energisystemet utvärderas i konkurrens med andra teknikval
(exempelvis gasmotor), innan större satsningar mot marknadsimplementering kan sjösättas. Det behö-
ver utredas under vilka processmässiga förutsättningar som elproduktionen blir så effektiv som möj-
ligt, och vad som i detta avseende skiljer biomassaeldad EFGT från utnyttjandet av andra processer
(exempelvis VOC, metan och biogas). I litteraturen finns ett flertal teoretiska studier kring systemin-
tegration av EFGT, både för biomassaförbränning och för andra källor, exempelvis [2-12]. Emellertid
är informationen relativt fraktionerad och svårtillgänglig för direkt tillämpning i en industriell appli-
kation. Utöver detta saknas ekonomiska bedömningar av tänkbara systemalternativ. För att svensk in-
dustri skall kunna tillgodogöra sig den existerande kunskapen och självständigt värdera den, krävs där-
för att den tillgängliga informationen sammanfattas och kompletteras, samt att ekonomiska potential-
bedömningar görs. Vidare är den totala potentialen en nyckelfaktor, samt vilka typer av anläggningar
och applikationer som är bäst lämpade för externeldad gasturbin och andra teknikval. I detta saman-
hang är potentialstudien av Bernotat och Sandströms [13] en god utgångspunkt för att komplettera de
källor som idag finns och som skulle kunna utnyttjas för EFGT och andra tänkbara teknikval, samt
7
vilka framtida nyexploateringar som kan förväntas. Externeldade gasturbinapplikationer är emellertid
inte begränsat till biomassaeldning. Slip-strömmar av metan och VOC från industriella processer,
gruvindustri samt bio- och deponigas är andra potentiella områden. Dessa gaser är, liksom bio-
massarökgaser, generellt alltför förorenade för att elda direkt i en gasturbin. Här skulle EFGT och ex-
empelvis gasmotorer kunna bidra till ett tillskott till den svenska elproduktionen samt till utveckling av
nya tekniknischer för svensk industri. Emellertid krävs en fördjupad översyn av vilka potentialer och
utmaningar som finns. De olika teknikvalvens nisch i det svenska energisystemet behövs dessutom ut-
värderas mer ingående, innan större satsningar mot marknadsimplementering kan sjösättas.
Implementering av EFGT på marknaden har, förutom marknadsosäkerheter, hittills hållits tillbaka av
tekniska utmaningar. Dessa är främst kopplade till utvecklingssprång inom:
Värmeväxling,
turbinutveckling.
Turbinutveckling är ett central och mycket viktig utvecklingsarbete. Ett fungerande kommersiellt
gångbart turbinsystem är ett måste för att EFGT skall kunna realiseras på marknaden. Mikroturbiner
för applicering mot EFGT har emellertid varit svårt att finna på marknaden. Emellertid har på senare
tid gjorts viktiga framsteg. Bland annat arbetar det svenska företaget Pomero AB/Ecergy AB, med att
utveckla en turbin i storleksklassen 100 kW specifikt för förutsättningarna vid externeldad gasturbin.
Vidare har Siemens visat intresse för marknaden kring teknik för energiåtervinning av slip-gaser i Au-
stralien. För att en marknadsmässig turbin för EFGT applikationer skall kunna utvecklas behövs emel-
lertid fortsatt dedikerad forskning och utvecklingen inom området, med fokus på säkra och användar-
vänliga styrsystem. I detta projektet har därför ett starkt delfokus varit inom just det området.
Utmaningen i värmeväxling mellan turbincykel och eldstad ligger i att materialmässigt klara den krä-
vande miljön i de heta rökgaserna från en bioeldad närvärmepanna, kombinerat med att kunna designa
en värmeväxlare som är ekonomiskt gångbar. För att elverkningsgraden ska bli tillräckligt högt krävs
att den komprimerade luftens temperatur i turbincykeln når minst 750 °C, vilket motsvarar ca 850 C
på rökgassidan. Detta medför utmaningar materialmässigt, även i rena gaser. Dessa kan till viss del lö-
sas genom att använda sig av temperaturtåliga legeringar, men detta leder lätt till höga totalkostnader. I
applikationer med förbränning av biomassa eller starkt förorenade VOC-strömmar, adderas utmaning-
ar i form av beläggningsbildning och korrosion på de värmeöverförande ytorna. Vid förbränning av
skoglig biomassa är det framför allt kondenserbara gaser och partiklar av alkali (K+Na) som tillsam-
mans med svavel, klor och koldioxid bildar olika typer av saltbeläggningar på värmeväxlarytor. Detta
kan leda till direkt nedsättning av verkningsgraden, reducerad tillgänglighet samt allvarliga materialin-
teraktioner och korrosion. Problemet har tidigare framför allt försökt lösas genom att skapa en pro-
duktgas med lägre halter av dessa oorganiska ämnen, genom att använda sig av specialdesignade för-
bränningssystem eller förgasning istället för direkt förbränning [3-6]. Uppfattningen är emellertid att
varken förgasning eller specialdesignade EFGT-förbränningssystem är i dagsläget kommersiellt gång-
bara för närvärmeskalan. Kraftvärmen bör produceras i en konventionell anläggning, under specifikt
anpassade driftförhållanden, för att bli ekonomiskt lönsam. Därmed kvarstår högtemperaturvärmeväx-
ling i förbränningsrökgaser som en rimlig lösning på kort- och medellång sikt, för att EFGT skall
kunna bli ett attraktivt marknadsalternativ. I det här projektet har därför en riktad satsning gjorts på
tillämpad forskning kring beläggningsbildning (rökgas-vvx interaktioner), materialfrågor och värme-
växlarutformning för EFGT.
2 Mål
Den övergripande målsättningen med projektet var att generera kunskap och tekniska underlag som är
direkt tillämpbara för utvecklingen av EFGT applicerat på biomassaförbränning och olika processer
och bränslen i det svenska energisystemet.
Specifika mål med projektet var att:
8
Kartlägga marknadsmässiga och ekonomiska potentialen för EFGT från biomassaförbränning
och andra processer, såsom destruktion av flyktiga organiska ämnen eller metan, jämfört med
andra tänkbara teknikval (exvis gasmotor)
Identifiera tekniskt och ekonomiskt fördelaktiga systemintegrationslösningar, vilka för de
olika processerna är applicerbara för värmeväxling och turbincykel för EFGT
Definiera kritiska parametrar och processmässiga strategier kopplat till beläggningsbildning
och korrosion på högtemperaturvärmeväxlare
Bestämma designkriterier (material och dimensionering) och driftstrategier för en värmeväx-
lare i rökgasapplikationer med biobränsleeldad rostpanna, med specifikt fokus på värmeför-
sörjning av externeldad gasturbin
Kartlägga behov för och ta fram nytt styrsystem och kontrollalgoritmer för turbindelen i exter-
neldat gasturbinsystem.
3 Genomförande
Arbetet i projektet har genomförts i tre parallella arbetspaket (AP):
AP 1 Systemstudier externeldad gasturbin
AP 2 Högtemperaturvärmeväxling i rökgaser från biobränslepanna
AP 3 Styrning av turbin samt integrering med befintlig panna
Nedan redovisas arbetet och resultaten från respektive AP i separata kapitel. Delar av materialet redo-
visas på engelska, dels på grund av att vissa medförfattare inte har svenska som modersmål och dels
på grund av att skrivandet på engelska syftat till utbildning i akademiskt författande för de doktorander
som medverkat i arbetet.
4 Systemstudier externeldad gasturbin
AP1 har utförts av Per-Olof Sjögren, doktorand vid Chalmers, och arbetet har handletts av Sven Her-
mansson, SP, och Mikael Odenberger, Chalmers. Stöd har givits från Enertech AB Osby Parca och
Ecergy AB i form av indata och tekniska specifikationer. AP1 har finansierats av Energimyndigheten
och MEGTEC. Här redovisas arbetet och dess resultat i sammanfattad form. Fullständig rapport finns
att tillgå via författaren.
4.1 Objectives and method
The aims and objectives of WP1goals are:
o System integration studies - identify the beneficial integration solutions applicable for the heat
exchanger and the turbine cycle (EFGT).
o Map the potential for EFGT – identify market and the economic potential of the EFGT from
biomass combustion and other processes, as organic compounds and methane compared with
other technology choices.
o In conjunction with WP2 and 3, identify design criteria, critical parameters, operating and
process related strategies linked to corrosion and fouling for the high temperature heat ex-
changer operating in biofuel grate boiler flue gas.
The main focus in AP1 has been to evaluate the implementation of an EFGT system on existing fur-
nace constructions for heat only producers. It is supposed to be implemented with relatively small
modifications and with focus on fuel efficiency and moderate investments. The aim is to produce
power at reasonable costs and low additional fuel consumption where no other more economical alter-
9
native exists today. An evaluation niche for this system approach is the district heating (DH) systems
in Sweden, where a comprehensive set of data exists. For a high utilisation potential, DH-base load
heat only boilers operating on wood fuel in a suitable size range have been used for the evaluation.
With limiting the fuel choice to wood origin may also give a better chance to standardise the high
temperature heat exchanger materials. Based on above criteria, the EFGT systems base load potential
with related efficiencies is evaluated. Also the added power generation potential by dual firing or in-
creased humidification have been simulated, their influence on the fuel efficiency and enhanced power
output calculated and load variability advantages discussed.
4.2 Small-scale CHP cycles
The conventional steam cycle based combined heat and power (CHP) plants are typically cost effec-
tive down to about 20 MW thermal. Smaller plant sizes are to the majority heat only producers due to
the difficulty to justify the increased investment cost and low or reduced electrical efficiencies normal-
ly related to small-scale generation systems.
Among technologies that have been proposed are:
Steam based down sized and simplified concepts like using close to saturated steam with a
simplified steam turbine [14], or with saturated steam and a screw expander [15]. These are
typically designed for thermal powers of > 5 MW. Reciprocating steam engines can be used
for thermal load in the 1 to 10 MW thermal range and supply 100 – 1000 kW e.
Organic Rankine Cycle (ORC) uses a refrigerant type of media (for example HFC’s) for ex-
pansion and condensation at moderate temperature levels. Typically, the heat is extracted with
a hot oil system operating at low pressure and temperature level which reduces design and ma-
terial issues. The technology has been widely applied (several hundreds of plants) for electrici-
ty generation in a range from 200 kW to 10 MWe with the optimal application range in 500
kW to about 2 MWe [16, 17, 18]. Smaller systems are approaching the market as for example
based on screw expanders in the 50 to 500 kW range [19] and a reciprocating engine based
micro system “Craftengine” generating 2 – 10 kW e per unit [20].
Stirling engines typically require a high temperature level in contact with the engine to obtain
a good efficiency. For direct heat transfer this is a challenge material vice and another ap-
proach to obtain the required temperature level is to use a gasifier and combustion chamber.
Practically this increases complexity and considering the relatively small unit sizes of Stirling
engines, typically from single kW’s to 35 kW e it becomes cost intensive per kW.
Externally fired gas turbines (EFGT) using heat for the compressed air instead of a conven-
tional combustion chamber. The heat transfer from the furnace gas to the compressed air is ob-
tained by a high temperature heat exchanger and the turbine airflow is totally isolated from the
flue gas, so the aim is to keep the turbine in clean air for a minimum of wear. A number of
theoretical studies have tested this approach thermodynamically and presented very promising
results [21, 22]. Other studies have found a promising application area for the technology,
considering the reasonable cost level of small scale gas fired turbine systems in the 50 to 200
kW range, or example [23]). Pilot units have to some extent confirmed the studies in short
time tests, but none of the biomass converted pilot units has, to our knowledge in the group,
had any long track record for proving economic viability.
10
Fuel conversion technologies like gasification for direct use in a combustion engine is to some extent
addressed in the economical comparison part of the project, but gasification and conversion to biofuels
have not been included in this study.
4.3 The modelled EFTG system
A short clarification of the three item numbers among the components in Figure 2.
1) The humidifier that reduces the temperature after the compression and makes it possible to
transfer more heat from the flue gas and it is also adding mass and volume flow to the turbine
for increased power-
2) To use parts or all of the turbine exhaust air as combustion air. It can either be done directly
after the recuperator or, if the temperature is too high, after cooling by the water circuit.
3) Heat recovery by pre heating of return water before furnace.
The triple line arrows in the picture below represents both the boiler air and flue gas flows and the tur-
bine air flow after heat uptake from the flue gas.
Figure 2. Flow sheet of the EFGT process
Osby Parca boiler
The furnace data used is based on an Osby-Parca boiler type PB2, illustrated in Figure 3. Osby-Parca
boiler type PBFel! Hittar inte referenskälla., suitable for up to 3 MW thermal load and is fuelled with
pellets of 10% humidity. The excess heat is distributed to water around the furnace, by double-sided
water filled walls and to flue gas cooling downstream the combustion zone. The efficiency is claimed
to be 90.2% (at 130 0C flue gas exhaust temperature).
11
The Osby-Parca boiler is here used as a model of existing district heating (DH) and industrial hot wa-
ter production units, in a common size range (2 – 5 MW), with potential use as heat source for an add-
ed EFGT system. If the demo integration succeeds, it is likely to prepare future furnaces for more easi-
ly and effectively connect such an option or adder. The modelling also explores system possibilities
with more humid wood fuels, for comparison purposes of fuel and humidity influence, the typical
boiler data has been kept for comparison purposes, even if the PB2 boiler is not made for 50 -55% wet
fuel.
Turbec T100 module – description.
The T100 Turbec module consists of the compressor and turbine combined with a recuperator, which
is an internal gas/gas-heat exchanger, preheating the compressed air with the hot air at the turbine out-
let. The combustion chamber (C.C) is to be replaced with connections to an external heat exchanger
(HTHE). The module includes power electronics to force start the turbine and then when heat and
pressure is in excess, produce power through the generator with AC conversion electronics for grid
distribution. The power section of the module is designed for 100 kW e generation at a turbine inlet
temperature (TIT) of 950 0C. A reduction of the TIT to 750
0C will typically give a power decrease of
30 to 40%. To utilize the power section potential at 750 0C TIT, a modified compressor and turbine
would be needed. Theoretically this would typically be done with a lower compression ratio and high-
er mass flow (for optimum efficiency). This design change has not been implemented yet. But there
are other means to boost the power and utilise the potential of the power section. Two options have
been modelled for evaluation. Either by humidification after the compressor, or by co-firing a second
fuel after the HTHE heat up. Theoretically, both can be combined and boost the power above the exist-
ing power electronics capacity or be used as power moderators (adders) at boiler part loads. The mod-
ule also includes ventilation equipment that contributes to the Turbec modules internal consumers
which reduces the net power out from the module. The reductions is approx. 7.5 kW at nominal power
(information from Ecergy) and for simulations a 8% loss of the generator electrical output has been in-
cluded. The heat losses from the turbine module are approximately 35 kW in standard operation with
the using a burning chamber and 950 0C TIT. At a reduced temperature level of 750
0C TIT the losses
are assumed to be 25 kW (for modelling purposes).
Figure 3. Osby-Parca boiler type PB
12
Figure 4. Turbec T100 microturbine
Figure 5. Turbec module in exploded view
High Temperature Heat Exchanger (HTHE)
The high temperature heat exchanger replaces the combustion chamber and transfer heat from the oxi-
dised flue gases to the compressed air after the recuperator to obtain a TIT of 750 0C. The flue gas
temperature to the heat exchanger has been set to 850 0C; maximum permissible pressure drops for the
clean- and flue gas- sides have been given (see the appendix section). Based on these values, heat ex-
changer geometry has been proposed by LTH and the detail data has been calculated. These have then
been used for model evaluations and for a cost estimate. The heat exchanger is designed to be exter-
nally positioned between the furnace and the turbine module, which gives flexibility for retrofits of ex-
isting plants. The EFGT system is at this stage more of a generalised “add on” design, for implementa-
tion on different furnace types and sizes. The input data, design results and a cost estimate can be
found in the appendix section.
A similar integration solution, but in a 10 time larger scale, technically can be found in the study from
Värmeforsk [23]; “Pre-study of an externally fired turbine in combination with a hot water furnace
for solid biomass”.
The system was evaluated for a biomass furnace of 20 MW and with an EFGT system delivering one
MW. One difference is that they have based the study on a turbine without recuperator. The heat loss
13
is avoided by using the turbine exhaust air as combustion air. The high temperature heat exchanger
(HTHE) is here proposed to be sectioned for different temperature levels (wide operation range with-
out recuperator). Materials in standard high temperature pressure coded steel qualities limits TIT to
750 0C. They have quite detailed explanations of how they modify the standard turbine and boiler, the
position of the heat exchanger and alike. The HTHE was designed and cost evaluated by a manufac-
turer and the calculated weight for the HTHE was totally about 35 tons where of about 30% were steel
tubes (10-11 tons). The temperature levels for the flue gas temperature, the TIT, the mass flow to
power relation and the resulting moderate electrical efficiency was similar to this project (in its closest
match of configurations).
Heat recovery of the turbine module exhaust air
The air temperature after the recuperator is typically on the 250 0C level and heat recovery of the air
stream is needed to reach high thermal efficiencies. For the evaluations, it is anticipated that a heat re-
covery module for the hot air after the turbine will be used to preheat the return water from the DH
system. The turbine exhaust air is here anticipated to be cooled to 90 0C to meet a return water of
about 70 to 80 0C. A lower DH system return temperature could potentially be used for low tempera-
ture applications resulting in higher thermal efficiency. Another way to increase the efficiency is to
use the outlet air as oxidation air in the furnace. With all the turbine flow going into the furnace, a sep-
arate stack loss is avoided and the turbine outlet air loss eliminated. However, existing furnaces may
not be easily adapted for supply of hot air as primary and secondary air. So also in this configuration it
could be needed to cool the turbine exhaust air with a water circuit, before using it as combustion air.
Humidification module
The humidification module uses the warm air after the compressor (about 200 0C) to evaporate liquid
water that expands and increases the volumetric flow into the turbine and thus increases the power
generated. The module can be made as a small scrubber, a vessel filled with distribution packing,
which evaporates the liquid in gently and controlled manner (this is the modelled version, with some
pressure drop penalty). The advantage with this evaporative type of humidification is that a lower
quality water can be used than by using atomisation nozzles, but it introduces an extra pressure drop.
A nozzle injection system will be cheaper and give a minimal pressure loss for the compressed air, but
will require higher water quality demands when impurities may get air borne and go through the tur-
bine. If the humidification is only used intermittently for power compensating purposes, this last ver-
sion may be the most cost efficient. In both of these two options there is very little pump work needed
to pressurise the water compared to the volume increase (about a factor 1500), that with the pressure
level adds force in the turbine. The T100 design has been evaluated with an adder of 7.2 % weight
H2O, which can be easily evaporated in all modelled versions. The power increase potential is substan-
tial (>30% to 40%), so the humidification module may substantially increase the power production in
a fixed Turbine geometry. It is interesting when looking at the investment cost to power potential and
can be used for boosting purposes, as a possibility to generate more electricity when the circumstances
are right or as a moderator, to stabilize the output power, when the furnace is operating at part loads.
On should be aware of that the heat input need from the furnace is substantially higher than the power
increase, this due to the energy penalty for vaporisation. If the energy in the water vapour is emitted to
atmosphere without condensation, the heat loss and reduction in thermal and marginal efficiency will
be substantial. If not the fuel is regarded as very cheap or the electricity spot price is very high, humid-
ification has to be combined with condensation for good total and marginal efficiencies.
Condensation module
If we are looking at constantly increased power production by humidification, there have been two dif-
ferent types of cooling modes modelled, this to regain the heat from the vaporised water. One is simp-
ly to cool the air to the 40 to 50 0C level to retain a large part of the humidity as condensate. Especially
14
the combination of using the humidified turbine air as combustion air together with humid wood fuel
gives a high recovery potential already at cooling to 50 0C. A low temperature water circuit can cover
needs as drying or be used in a low DH system (floor heating, DH generation 4 system or alike). As a
second possibility, it is possible to use a heat pump to cool the exhaust further. With an exit flue gas
temperature in the 10/15 0C range, even heat recovery by condensation of dry pellet fuel flue gases or
humidified turbine exhaust air can be done alone or in combination.
The continuously condensed water will keep concentrations of acidity and impurities at moderate lev-
els and the pre cooling by the normal water circuits; will also keep the temperature operation range for
the condensation section (inlet air/gas < 150 0C) at reasonable levels for material or coating selections
of the parts used.
For the efficiency potential estimate, there’s assumed a conventional compressor driven heat pump us-
ing 1 part of electricity for 4 parts of recovered heat. The electricity needed is internally generated by
the EFGT, which in the tables are shown as a decreasing el. efficiency and an increasing heat efficien-
cy.
Top up heat for increased TIT – Dual fuel combustion chamber
The T100 turbine is specified for 950 0C TIT, this without using extreme materials, blade cooling or
any typical high temperature turbine features. There is possibly a small cost reduction possible, proba-
bly very small in relation to total cost of implementation, if optimising the material quality to the low-
er temperature. On the other hand, a higher temperature level will give a higher power and efficiency
potential and the temperature can be increased to the natural gas TIT of 950 0C. The value is not only
the maximum generated power but also flexibility for generating power when needed at reduced boiler
duties. It may however not be possible to use a standard gas turbine burning chamber for this (now re-
ceiving up to 750 0C air after HTHE instead of about 600
0C after the recuperator in standard natural
gas burning mode). The values for the combustion adder in the simulations are related to methane,
which could come from biogas origin for fossil free operation.
15
4.4 Cost estimates
The modelled EFGT system
Table 1. Cost estimate for Turbec EFGT implementation on existing boilers
Cost estimate for EFGT applied Project T100 Modified T100
on existing boilers 100 kW gross
Total system el. output (net*) 64 90 kW
Turbec T100 1 500 000 1 500 000 SEK
HTHE 300 000 400 000 SEK
Adaption furnace with HT valve 250 000 300 000 SEK
Installation & start up 300 000 300 000 SEK
Total Cost est. 2 350 000 2 500 000 SEK
Cost / kW el. output (net) 36 700 27 800 SEK / kWe *) The net power is after reduction for the T100 modules internal consumers and the flue gas fan power adder.
In Table 1 fundamental indata for the cost estimates are given. The 64 kWe net version is based on the
integration of a standard T100 module without modification of the compressor/turbine. The 90 kWe
net version has an assumed modified compressor and turbine for increased mass flow to match the
generator and electronics potential of the turbine module. The modification is not assumed to increase
the serial production cost to ant significant extent. Some further explanation to the data in Table 1:
- Turbec T100 is the complete module for electricity generation and control of the turbine.
- HTHE will need to have a proportionally larger surface area for heat transfer in the modified case,
thus the cost increase. The material is based on a conventional high temperature steel with 1000 C
scaling temperature in air and low risk for grain growth/embrittlement in the HTHE operation temper-
ature range.
- Adaption to the furnace requires a high temperature (HT) -valve for creating resistance so the flue
gas passes the HTHE and will need connecting ducts and an air to water heat exchanger to recover
heat and cool the air before using it as combustion air.
- Even if the T100 system is containerised, there will be site assembly of the HTHE with ducts, some
power and signal cables to connect beside the boiler and start-up of the system.
16
Other comparable small scale CHP technologies
In Table 2. Investment comparison for small-scale biomass CHP technologies, cost estimates for other
comparable small scale CHP systems are given, based on literature values.
Table 2. Investment comparison for small-scale biomass CHP technologies
Technology Comment Size Power Total eff Marg.eff. Investm. Ref. Year
*) The increased heat loss at relates to producing water (with 90,2% eff.) Note 1: The marginal efficiency for the difference in additional heat need with and without humidification.
Table 7. EFGT exhaust to furnace - all air is utilised as the only combustion air source
*) The increased heat loss at relates to producing water (with 90,2% eff.) Note 1: The marginal efficiency for the difference in additional heat need with and without humidification.
With the turbine outlet flow fully integrated to the furnace, the average mass flow will be due to that
the fuel weight is added to the airflow and that. The marginal efficiencies then increase, due to a lower
power penalty for the furnace fan (less volume flow with the same assumed resistance) resulting in a
higher net el. generation. This could be compensated by another design for a lower HTHE flue gas
side pressure drop, but this will also result in a larger and costlier heat exchanger.
Using humid fuel instead of dry fuel
These system configurations are identical to previous sections, but the fuel is switched from 10% hu-
mid pellets to wood with 55% humidity. The wood composition is assumed constant so the fuel weight
of the humid wood and components in the wood has been reduced to exactly the half (now 45% dry
substance instead of 90%). The most important effect is then that the mass flow will increase in rela-
tion to heat generated and should, hence, result in a higher flue gas flow in the HTHE in relation to
fuel value and thus increase the power to heat ratio for the EFGT application. Another effect is that the
increased humidity makes condensation of the flue gas for a far driven heat recovery with means of a
low temperature system more feasible.
The simulations with 55% humid fuel, presented in Table 8, give similar marginal efficiencies but
lower heat and total efficiencies than with dry pellets. The reasons should be in that the higher fuel and
air mass flows compared to energy input gives a higher relative stack loss and the increased power to
heat ratio then “costs” in heat and total efficiency. The results indicate that a 1 MW furnace using wet
wood, will give enough mass flow for implementing a T100 module with exhaust to air and a 1,5 MW
furnace could be fully combustion air integrated to a T100 with higher resulting efficiencies and in-
creased power to heat ratio (compared to dry pellet operation). Based on fuel input the power genera-
tion potential is higher for all cases and for the most efficient configuration results in an increase of
about 20% compared to dry pellet operation.
23
Table 8. EFGT configurations with the furnace operating on 55% humid fuel
C. Dry-exh.to f. 1437 1131 111 63 4,4 83 51 Note: 1
D.Hum-exh.to f. 1437 945 297 87 6,1 72 26 11,9
Note 1: The marginal efficiency for the difference in additional heat need with and without humidification.
Note 2: The mass flow and fuel input of the furnace is decided by either using an enthalpy balance over the HTHE or by using the turbine outlet flow as combustion air.
Dual fuel mode – Support gas fuel added to the turbine cycle
The turbine is specified for 950 0C TIT. Therefore, it could be of value to increase or to modulate the
TIT from the design temperature of the HTHE (750 0C) and, thereby, increasing the power output
from the EFGT. Another important factor that can be improved by support firing is the annual utilisa-
tion of the power production unit. The utilisation of base load hot water producing units is typically 30
to 60% on a yearly basis (referred to the nominal load). Both the power and marginal efficiency gener-
ated by the EFGT module drops quickly with reduced TIT temperature, which is a likely scenario at
boiler part load. Therefore, a “boost” in temperature by support firing gas directly into the turbin cycle
could be used for power compensation and the possibility for added operational hours at nominal pow-
er may as well be of value even if a potentially more expensive fuel has to be used for the “make-up”
in power. Here, it has been considered how such method could contribute to the system efficiency. The
following cascade simulation cases have been used to evaluate this possibility:
1. Boiler part loads resulting in a TIT of 550 or 650 0C.
2. Addition of fuel to reach nominal level of TIT 750 0C.
3. Increase from 750 0C with added fuel to 850 or 950
0C for boosting the power output.
The simulation results of Options 1-3 are given in Table 9 and Table 10. The results show that the
power output of the EFGT can be increased by more than 50% by complementary firing from 750 to
950 0C. However, an added pressure loss from the added combustion chamber will reduce the net
power somewhat. Furthermore, the total fuel efficiency of the fuel adder is high. This is due to that the
stack loss is constant (always cooled to 130 0C in the simulation) and the increased heat loss due to the
temperature increase are low compared to the total heat load. This also means that the added fuel is
well utilized when replacing biomass; the total fuel marginal efficiency is increasing with the added
second fuel.
The flexibility with a second fast responding heat source will give the system a higher operative value.
It will also be possible to run the turbine system at low or no heat demand for increasing yearly power
production. The extra cost for the second fuel can to some extent be compensated by the relatively low
added operational cost for a slightly increased investment (compared to no second fuel system). On the
other hand, the results imply that the increased TIT also increases the turbine outlet temperature and
the recuperator increases the preheating of the compressed turbine air before heat exchange with the
flue gas. Therefore, the possible heat input, temperature difference, for the HTHE will be limited and
the biomass part of the input energy will be reduced with higher TIT’s. The biomass heat input would,
however, be more constant in a system without a recuperator (the increased turbine outlet temperature
at the higher TIT will not heat up the compressed air as with the recuperator), but the trade-off will be
that the HTHE has to operate in a larger temperature span, from a lower temperature, in the flue gas.
24
Table 9. Options 1 and 2, EFGT performance when using support gas firing to lift part load TIT to nominal TIT. Red text in-
dicates the support fuel cases. Supply fuel is methane (biogas). Turbine exhaust to Air.
Dual fuel eff. - T to air T bio. TIT by (3)
El. gen. El. eff. Tot.eff Tot. Fuel Fuel 2 Fuel 2
Efficiency net. values Boiler fuel input is 1349 kW
(0C) fuel2(
0C) net.kW net. % net.% m.eff.% el.eff.% tot eff.%
Furnace heat only - no tur-bine
NA NA 0 0 90 NA NA NA
Part load - no fuel adder 550 550 22 1,6 83 18 NA NA
Part load - no fuel adder 650 650 43 3,2 83 27 NA NA
EFGT: HTHE bal, - nom. case 750 750 64 4,7 82 34 NA NA
Fuel adder to nominal power
(1) 550 750 64 4,2 84 33 22 89
Fuel adder to nominal pow-er
(1) 650 750 64 4,4 83 33 22 90
Fuel adder for TIT increase (1)
750 850 86 5,9 83 38 22 90
Fuel adder for max.TIT (1)(2)
750 950 109 7,0 83 42 22 91
Table 10. Option 3, EFGT performance when using support gas firing to lift TIT above nominal TIT. Red text indicates the
support fuel cases. Supply fuel is methane (biogas). Turbine exhaust to Furnace.
Dual fuel eff. - T to furnace T bio. TIT (0C) El. gen. El.eff Tot.eff. Tot.fuel Fuel 2 Fuel 2
Efficiency net. values (0C)
fuel 2 (
0C)
net. kW net.% net. % m.eff.% el.eff.% tot eff.%
Furnace heat only - no tur-bine
NA NA 0 0 90 NA NA NA
EFGT: HTHE bal,-nom.case 750 750 63 3,7 87 50 NA NA
Dual fuel for TIT increase(1)
750 850 89 5,0 87 56 25 85
Dual fuel for max. TIT (1)(2)
750 950 119 6,2 87 59 25 85
Furnace flue gas recirculation – maximising the mass flow through the HTHE
A common assumption for EFGT systems is that the majority of the heat generated can be transferred
to the compressed air. But in practice, there are some obstacles for the heat transfer to consider:
Radiation heat transfer in the combustion zone can absorb a lot of heat before a more conventional
heat exchanger, but reactive particles (reducing atmosphere), slag, fluctuating temperatures make
material selection problematic/expensive, and protective material may limit the heat transfer. Heat
transfer gas to air requires large surface areas compared to heat transfer to water.
25
To create cooling surfaces to close to the combustion zone to obtain a higher flue gas temperature
into the HTHE is problematic; not only for the materials, but also for the emissions (forced cool-
ing too near the combustion zone may stop burn out of fuel, CO and increase soot).
For simplicity, the system can rely on heat transfer by the flue gas mass flow after complete oxida-
tion, which gives a stable atmosphere with a controlled oxygen amount, and possibility for re-
duced slag amounts.
However, the mass flow to heat release ratio for a furnace is small compared to the mass flow to
heat need for an EFGT system. This puts limits to heat transfer based on flue gas mass flow.
This projects EFGT configurations and simulations uses this last conservative (or robust) design which
limits the net electrical efficiency in the range 4 to 8 % based on biomass fuel value, even with bal-
anced mass, or balanced enthalpy flows in the HTHE. To exceed these values one has to take one or
several of the measures:
Radiative heat transfer surfaces “connected” to the compressed air.
Radiative and convective heat transfer in the combustion zone.
Flue gas recirculation to match the compressed air amount at lower boiler heat release.
The last option seems most feasible by means of controlled and modelled. The adder to the earlier
simulation models is a flue gas loop, which includes the recirculated gas’ pressure drop penalty
through the HTHE for the combustion fan.
There is one commercial system that seem to have been custom designed based on this recirculation
approach and that is the “Schmid moving grate UTSR 1200 & EFGT” [28]. I have not found any detail
data about this unit more than Schmid’s data sheet, but it seems to lack the recuperator so the heat ex-
changer would then operate with flue gases from 900 down to 3000C. The turbine air would then be
heated up from 200 (after compression) to 750 0C. They claim to reach 15% electrical efficiency.
For more data on the Schmid unit, see the appendix section.
The recuperator, in an EFGT configuration, operates in almost clean air compared to the HTHE. In
addition, the recuperator increases the efficiency of a gas turbine system, especially in power only
mode where it reduces the exhaust temperature (stack loss) of the system considerably. If the system
operates in CHP mode where there is always need for hot water, the hot exhaust gases from the turbine
can be used for hot water production (with a recuperator or not). In the case of the EFGT application,
the electrical efficiency and power to heat ratio can be increased, without a recuperator, due to that
more heat (a wider temperature range) can be transferred in the HTHE. However, a wider temperature
range in the flue gas for the HTHE may be more challenging from a construction and contamination
(slag) point of view.
To determine the increased el. efficiency potential for the recirculation option, following models were
simulated.
- The Schmid model. Based on the data found a model of the system was run and evaluated in
EBSILON. Schmid uses 50% wet wood fuel (also used in the simulation).
- The Turbec T100 system with the normal data but without the recuperator. The HTHE was ex-
tended downwards in temperature range but with the same supply (850 0C) and minimum dif-
ferential (100 0C) temperature. With 50% wet wood fuel.
- The T100 system as above but with recuperator as the standard system and 10% humid pellets
(for comparison to earlier evaluated models).
The recirculated amount of gas has been set for balancing the mass flows. The recirculated amount is
typically a factor of nearly 2 times the need for combustion air so the flow through the HTHE is al-
most 3 times the amount without recirculation based on the same fuel heat release.
26
Table 11. EFGT models with flue gas recirculation
Data / Item SCHMID SCHMID T100 data Turbec config Unit
/ Efficiency Claimed Simulated Schmid config. and data
Marginal eff. Net Not given 44,6 44,8 43,8 % *) Not known if the value is a real net including all internal consumers or if it only includes the flue gas fan.
**) First value with the turbines outlet air is cooled to 70 0C, // second value, if cooled to 90 0C.
The results are given in Table 11. The net electrical efficiency has a potential to be more than doubled
with flue gas recirculation, but from a total efficiency point of view, there is one drawback. Only about
a third of the turbine outlet flow can be used as combustion air, the rest has to be cooled separately and
the total efficiency depends largely to the cooling temperature and the total stack losses. Even with
cooling of the turbine exhaust air to 70 0C, as in the examples above, the total efficiency is about 10%
lower than the best-integrated simulations found in earlier, and cooling to 90 0C, more in line with ear-
lier examples, would give a further reduction of 2%.
Heat recovery potential by flue gas condensation
It has been demonstrated that humidification after the compressor can maintain the turbine module de-
sign power when decreasing the TIT from 950 0C to 750
0C. This to a reduced total efficiency and to a
very low marginal efficiency. The power increase by humidification is however substantial (about +35
to 40% net) and could be used as a fast way to moderate the power output, either for retaining power at
reduced furnace load or for electricity generation increase. However, also reach acceptable total effi-
ciency the latent heat of the fuel gas moisture has to be. This can be achieved in a flue gas condenser,
which in this project has been investigated by the following options:
Flue gas condenser with a water circuit of 45 0C DH return temperature
Adding a heat pump to lower condensation temperature to 15 0C. The heat pump is applied af-
ter the passive system to deliver heat on the same level. COP of the heat pump is assumed to
be 4 at an upper delivery temperature of 50 0C.
The results of the thermodynamic system simulations are given in . EFGT net output power can be
seen to increase with up to 40%, with potentially high marginal efficiency (> 60%). Furthermore, the
thermal and total efficiency is significantly increased. The positive results are achieved with firing wet
fuel or using humidification of the turbine cycle, and especially when combing the two. When using
dry pellets and no humidification of the turbine cycle (Case A-E), the condensation temperature of the
flue gases will be around 50 0C. There is, hence, no surplus of cooling with a 45
0C media. However,
with humidification of the turbine cycle (Case D), the condensation temperature will be higher than 50 0C. Condensing against a heat sink of 45
0C could then be favorable. The same goes for when using a
wet fuel of 55 % (Case F and H). The results show that the heat recovered with this direct cooling sys-
27
tem is ranging from about 10% up to 30% of the fuel heating value, which results in considerable in-
crease marginal and total efficiency. The power generation is only slightly decreased due to moderate-
ly increased pressure drops and water circuit pump needs.
Even if condensing with a heat sink of 45 0C there is still a significant amount of water vapor left in
the flue gas – the residing latent energy represents more than 10% of the input fuel value. By using a
recovery system with a heat pump, the majority of the latent heat in the flue gas humidity can be re-
covered. With the heat pump in operation, the heat generation and total efficiencies are further in-
creased, according to the simulation results up to 115 % in total efficiency. However, the downside is
a decreased net power generation from the EFGT- system caused by the increased internal use of
compressor power. If there is use for low temperature heat, the EFGT with humidification may act as
the power source for both the heat pump and the flue gas fan when still delivering a high thermal effi-
ciency.
The results, which are presented in Table 12, show that the system this humidified turbine system has
far higher efficiencies in the condensing modes and that the heat is especially easy to recover when fir-
ing a humid fuel. The humidification adder will then be recovered at relatively high condensation tem-
perature. The extra humidity that is added to the turbine system does not only increase the net power
output from the system from 22 to 25 kW, it also delivers this with a heat penalty of only 29 to 39 kW,
equivalent to 71 and 60% marginal efficiency respectively (compared to the earlier estimated marginal
efficiencies of 12 to 13 % in the non-condensing modes).
Table 12. EFTG with flue gas heat recovery, including options with heat pumping technology.
EFGT configuration & Fuel Furn.heat Cond.heat (kW) Tot.heat Net P**
El.eff. Tot.eff.
With 1694 kW fuel (LHV) 10% hum. water (kW)
50O
C 15O
C kW* kWe % %
A Base mode. Dry pellets 1439 63 3,7 89
B Base mode. Dry pellets & Heatpump 1439 0 248 1662 1 0,1 98
C Base mode. Dry p. with humid. oper. 1252 1227 88 5,2 78
D Base m. Dry p. hum. oper.&cond. 1252 209 1436 88 5,2 90
E Base m. Dry pellets humid operation condensation & heat pump
1252 209 195 1631 39 2,3 99
With 1436 kW fuel (LHV) 55% hum.
F Wet fuel, with condensation 1154 332 1461 63 4,4 106
G Wet fuel, with cond. & heat pump 1154 332 201 1662 12 0,9 117
H Wet fuel, with condensation 968 479 1422 87 6,1 105
I Wet fuel, with cond.& heat pump 968 479 211 1633 34 2,4 116
*) In the total heat is included a heat loss of the turbine module of 25 kW **) The internal generated el. is used for the heat pump.
28
4.6 Sensitivity analysis of component efficiencies
The following chapter investigates the sensitivity of the entire EFGT efficiency to changes in the per-
formance of included components.
Compressor, turbine and recuperator efficiencies
Figure 7. Compressor efficiency influence on net power production, at constant turbine efficiency of 87%. Flue gas to turbine
air heat transfer is constant at 286 kW
The compressor and turbine are specified for 77.6 and 87.3 % efficiency, respectively. However, Fig-
ure 7. Compressor efficiency influence on net power production, at constant turbine efficiency of 87%.
Flue gas to turbine air heat transfer is constant at 286 kWindicates that the electrical efficiency is
strongly dependent on the compressor performance. Therefore, the energy loss of the compressor may
be worth to consider as an improvement possibility. For example, a change from 77.6 to 85 % effi-
ciency for the compressor could give 18% in net power increase (from 74 kW 87.5 kW). The reason
for this high potential is that the net power out is the difference between turbine power generated and
the compressor power requirement. For example, if the compressor uses 225 kW and the turbine gen-
erates 300 kW, the positive result in generated power is 75 kW. Any percentage change in compressor
efficiency will affect the difference several times more, calculated in percent. There may, however, be
challenging to obtain such high thermodynamic efficiency, as 85% from a small compressor. It may
be more achievable for larger turbine systems.
Considering the recuperator, a high efficiency is especially valuable in power only mode. The heat loss
after the turbine outlet then becomes critical. When running the cycle CHP operation, the turbine out-
let air can be heat exchanged with the water circuit, meaning that the overall efficiency will compen-
sate for the loss in the turbine section.
Pressure drop in heat exchanger
A critical design factor is the HTHE pressure drop. The pressure drop on the compressed air side is not
critical. However, the pressure drop on the flue gas side can become considerable. Figure 8 illustrates
The sensitivity to pressure drop on the flue gas side is very high, due to the large volume flow through
the HTHE that has to be extracted by the boiler’s flue gas fan. Already a 1% (3.4 kPa) pressure drop,
related to the pressure increase in the turbine system (about 340 kPa), a power loss of 5% is achieved.
If applied on a larger furnace, the relative mass flow of flue gas will increase and the power loss will
become proportionally higher. Altogether, this means that pressure drop of the heat exchanger will
become an important parameter for evaluating the performance to cost basis when designing the
HTHE.
Figure 8. Power loss due to HTHE pressure drops. The mass flows are balanced, corresponding to a boiler of
about 1.5 MW (thermal) integrated with a 100 kW turbine module.
Turbine part load performance
The data curve from Turbec is included in the appendix section.
However not directly applicable in this project it is informative to see that the power characteristics is
optimised around 90 kW output and that the correction factor for the electrical efficiency only varies
from 1,00 +/- 0,01 in the 70 to 100 kW range.
In the simulations with EBSILON, the machine efficiencies are constant, which can be a reasonable
assumption within a narrow range with the compression ratio optimised for the new TIT, pressure ratio
and mass flow. The influence this would have on the simulations results should then be relatively
small. Also for power variation purposes, it is positive that the electrical efficiency is only moderately
reduced at 60 or even 50% output. The total efficiency correction relates to the Turbec CHP module
and is less informative for this project.
Turbine inlet temperature
In this project the turbine and HTHE is designed for a TIT of 750 0C. A demo or pilot plan unit could,
on the other hand, practically benefit from a flexibility to increase or decrease this parameter. This
will, however, affect the efficiency of the system. To illustrate the effect of such action, the influence
of TIT change on shaft power at a fixed mass flow has been simulated and the results are given in
Figure 9. Turbine Inlet Temperature (TIT) influence on power and efficiency. The power output is
30
seen to increases with increasing temperature. Lowering the temperature has the opposite effect. This
is related to two fundamental factors. With increased TIT, more heat is transferred into the turbine sys-
tem at a given mass flow. Then the efficiency rises due to the fact that the system is 1) optimised for a
higher TIT, and that 2) thermodynamically the efficiency will increase with increased TIT, given an
optimised pressure ratio. The summarized effect of the TIT on the power output and, thereby, the total
efficiency of the system is, therefore, a significant factor that should be included in the total cost-
benefit calculation of the system.
Figure 9. Turbine Inlet Temperature (TIT) influence on power and efficiency
Optimum turbine pressure ratio in relation to the TIT.
The system efficiency depends on the compressor pressure ratio and for a given turbine system there is
an optimum pressure ratio related to each TIT. Using a lower compression ratio than the optimal, a
higher mass flow has to be used to achieve the same power output. On the other hand, a too high com-
pression ratios for a given TIT level will result in an efficiency drop, and at a certain point the generat-
ed power will decrease. The power output will be reduced due to the fact that the compressor is pro-
ducing increasingly warmer air, finally approaching the limit of the recuperator operation range. How-
ever, a more efficient compressor gives room for a more increased pressure ratio.
The Turbec system was originally optimized for 950 0C TIT. This means that the design compressor
pressure ratio of about 4.4:1 logically is not the optimum ratio at 750 0C. An evaluation results instead
in an optimum compression ratio, of something between 2.5:1 and 2.7:1. As a comparison, the Schmid
EFGT system [28] has specified a lower pressure ratio of 3.8:1 and is then likely (detail component
data is not known) to be better positioned from an efficiency point of view at 750 0C TIT. It is there-
fore most possible that the EFGT system has a potential for improvement in efficiency, if the pressure
ratio is reduced, however keeping in mind that this might also affect negatively the power generated.
A fair compromise could be to modify the system for a pressure ratio of 3.5:1. The thermodynamic ef-
ficiency would then increase with about 2% (from a level of 26%). Because of pressure ratio reduc-
tion, the mass flow increase for the same power production would increase with about 5%.
31
A lower pressure ratio also puts less force to the hot pressurized parts so that is a slight advantage for
the total construction. What the HTHE is concerned, it will get a lower need for heat transfer (for a
given power) due to the higher efficiency, but at the same time, a slight mass flow increase is needed.
So the influence here cost wise is probably insignificant (with this examples limited pressure change).
Heat loss and internal electricity consumption in the turbine module.
The anticipated heat loss from the turbine, recuperator and connecting ducts at TIT 950 0C and 750
0C
is 35 kW and 25 kW respectively. This reduces the theoretical marginal efficiency 750 0C with ap-
proximately 20 %. There is, hence, substantial room for improvement; the surface areas are not so
large (except for the HTHE) to insulate with microporous “super” insulation or to run a return water
circuit pipe through a double walled insulation design that can be at 200 0C or more on the inside. In
addition, the internal electricity use within the turbine package is approximately 7.5 kW of the turbine
generated power (107.5 kW gross gives 100 kW e net output). This corresponds to 7% of the produced
electricity.
Outcome from the sensitivity analysis in respect to modification potential.
The comments are related to the parameters and simulation models for the T100 turbine module. To
increase the compressor efficiency gives a significant power generation increase due to that it reduces
the temperature after the compressor, and a second effect is that the optimum pressure ratio can be in-
creased which both is related to a smaller volume flow (less resistance per mass fraction in a given ge-
ometry) and higher thermodynamic efficiency.
TIT increase has a high impact on possible power production at a given mass flow.
The recuperator efficiency has an effect on the HTHE operation range and the possible heat transfer
from the flue gases. A highly efficient recuperator is beneficial if there is no heat production, but it re-
duces the possible heat-transfer from the flue gases to the turbine due to that it is limiting the tempera-
ture transfer range. The recuperator design/strategy should be evaluated based on application and
HTHE construction.
Reducing the compression ratio will give a higher marginal efficiency to a slightly higher mass flow.
However, to reduce the compression ratio all the way to the theoretical optimum may be counterpro-
ductive system wise, due to the reduced power to heat ratio, due to the turbine system reduced power
output in relation to the mass flow.
Heat loss reduction from the turbine system may give the largest single item potential for marginal ef-
ficiency increase. There is a potential for 10 to 20% marginal efficiency increase by reducing the heat
dissipation around the turbine components.
Part load performance in the 50 – 100% is not reducing the efficiency significantly and the 70 to 100%
range is very efficient (“compared to the sweet spot”) which is a positive factor if load changes or de-
liberate power modulations has to be accounted for.
The pressure drop in the system and especially in the HTHE on the low pressure, high volume, flue
gas side has to be considered so not too high parasitic losses is established due to flow restrictions. Es-
pecially if applying a small turbine on a large boiler. When the power module generates a small
amount of power in relation to the flue gas flow, a significant portion of the now limited gross power
generated will have to be used to overcome the extra flow restrictions on the flue gas fan.
The TIT is also critical for the performance. Increase in TIT has a large impact on possible power to
heat values. However, this does not come free, the HTHE material choices has to be affordable and
robust for operation in the temperature level chosen.
Flue gas recirculation increases the electrical efficiency and power to heat potential, but if reducing the
boiler combustion air under the turbine flow, some air has to be emitted in a separate stack, which re-
duces marginal and total efficiencies.
32
4.7 System power potentials
The potential of the EFGT concept for the Swedish market was evaluated for district heating plants
and industrial furnaces. For the district heating system a comprehensive set of data was used from
Svensk Fjärrvärme, which has been further developed by Naturvårdsverket to evaluate the market sec-
tor costs for new emission regulations [29]. During this project, the potential for implementation in the
industrial sector has, however, not been investigated independently, since no database for such instal-
lations have been found. As an assumption, the industrial potential as been approximated as two times
the district heating sector.
As comparison to the Swedish market potential, of comparable climate, fuel range and size, Austria
was selected [30] and, furthermore, other application areas than biomass combustion have been briefly
investigated.
4.7.1 Wood fired base load boilers in Sweden.
The potential for EFGT in the Swedish energy system (DH + industry) was evaluated from three dif-
ferent technological setups, identified within this study:
1) EFGT dry: 10% humid pellets; turbine integrated to the boilers with exhaust air cooled and fed
in as combustion air (3.7% electricity (to boiler fuel) at 52% marginal efficiency and 87% to-
tal efficiency).
2) EFGT humid: As 1) plus a humidification module to increase power. The extra power, in-
creasing from 3.7 to 5.2 % based on fuel (in kW power with about +40%) can be used for var-
iability, but is not fuel efficient, 27% marginal & 77% total efficiencies, unless combined with
flue gas condensation. With flue gas condensation the marginal efficiency may reach > 60%
and the total efficiency become close to 100%.
3) EFGT recirc: Flue gas recirculation with humid fuel (50%), but no humidification. Part of the
turbine exhaust is used as combustion air. The power generated can reach 15%. Without recu-
perator, which achieves 14,4% power to fuel and approximately 44 % marginal and 73 % total
efficiency. (The turbine part air that is not used a combustion air is cooled to 90 0C).
Common assumptions in all three cases:
- Each case is fully utilized for the size of boiler it is integrated with.
- All powers and efficiencies are net values with heat and internal power losses including the
extra power need for the flue gas fan.
- One full load year is assumed to be 4000 full load hours.
The summarized potential for the investigated technology options is given in Table 13 and Table 14.
The power potential is found to be fair – between 99 to 385 MWe (396 – 1538 GWhe) for the small
scale furnaces (1-10 MWth) depending on technological setup. Including also the medium scale fur-
naces (10-40 MWth), another 69 – 269 MWe (275 – 1071 GWhe) could be established.
33
Table 13. Power potential (MWe) for EFGT systems applied on boilers in Sweden
Installed boilers Base load Total Heat EFGT - dry EFGT - humid EFGT – recirc
in range (MW) units* (MW) MW e *** MW e *** MW e****
DH 1 - 5 199 458 17 24 66
DH 5 - 10 83 611 23 32 87
DH* total: 1 to 10 MW 282 1069 40 56 154
Industrial sector** 1 - 10 MW 423 1604 59 83 230
Total est. 1 to 10 MW 705 2673 99 139 385
DH total 10 - 40 MW 44 744 28 39 108
Industrial sector** 10 - 40 MW 66 1116 41 58 160
Total est. 10 to 40 MW 110 1860 69 97 269
Total est. all 1 to 40 MW 815 4533 168 236 655
* District heat (DH) Furnaces of base load type fired with wood fuels.
** The amount of heat generated by industrial furnaces is likely to be two times the DH (1,5 x DH used).
*** Based on the config. with the turbine exhaust air used as comb. air. Gives low P/H ratio but high marginal efficiency.
**** Furnace with recirculated flue gas for increasing the mass transfer to the heat exchanger and increase P/H ratio.
Table 14. Generation potential (GWhe) for EFGT on 1 to 40 MW wood fired boilers in Sweden.
Units in range - installed Production EFGT /y Production EFGT /y Producrion EFGT /y
capacities in ranges (MW) Dry in GWh e * Humid. in GWh e * Recirc. in GWh e
DH 1 - 5 68 95 263
DH 5 - 10 91 123 341
DH total: 1 to 10 MW 158 222 615
Industrial sector** 1 - 10 MW 237 333 923
Total est. 1 to 10 MW 396 556 1538
DH total 10 - 40 MW 110 155 429
Industrial sector** 10 - 40 MW 165 232 642
Total est. 10 to 40 MW 275 387 1071
Total est. all 1 to 40 MW 670 942 2609
*) The estimate assumes an average of 4000 operational hours at nominal power. **) Including up to double amount of heat generated by industrial furnaces, here assumed to be least 1,5 x DH.
34
4.7.2 Wood fired boilers in Austria
The results presented in Table 15 show that the stock of biomass boilers in Austria for the 1 – 10 MW
size range is almost identical to the Swedish stock in total installed heat. However, the average size is
somewhat smaller (about 3MW instead of 4 MW). This results in a potential in the same range as the
Swedish market, however somewhat smaller.
Table 15. Power potential for EFGT systems applied on boilers in Austria. Includes all furnaces both DH and industrial