Docstoc
EXCLUSIVE OFFER FOR DOCSTOC USERS
Try the all-new QuickBooks Online for FREE.  No credit card required.

TAR DEWPOINT ANALYSER

Document Sample
TAR DEWPOINT ANALYSER Powered By Docstoc
					ECN-C--05-026




            TAR DEWPOINT ANALYSER
   For application in biomass gasification product gases
                           S.V.B. van Paasen (ECN)
                             H. Boerrigter (ECN)
                               J. Kuipers (ECN)
                     A.M.V. Stokes (Michell Instruments)
                       F. Struijk (Michell Instruments)
                      A. Scheffer (Michell Instruments)




                                     Revisions
 A
 B
 Made by:

                           Approved & Issued:

 S.V.B. van Paasen
                                                           ECN Biomass
 Checked by:

                           H.J. Veringa
 J. Beesteheerde


                                                                     MAY 2005
Acknowledgement/Preface
This report describes the results obtained in the project “Tar dewpoint analyser”, conducted by
order of Novem under contract number 2020-02-12-14-010. The project has been executed by
ECN and Michell Instruments. Gerard Broers (ECN), Ruud Wilberink (ECN), and
Herman Bodenstaff (ECN) are acknowledged for their contribution to the project.


Abstract
This project aims at the development of an analyser for the on-line measurement of tar
dewpoints in biomass product gases. The work has been executed according to the project
proposal. On basis of the specifications for the tar dewpoint analyser (TDA), an existing
hydrocarbon dewpoint sensor was modified and a gas conditioning section was designed for tar
dewpoint measurements. Preliminary laboratory tests with the gas conditioning section and
dewpoint sensor were run to investigate the performance and fouling characteristics of the
dewpoint sensor and the gas conditioning section. The TDA (gas conditioning section + sensor)
was tested and validated downstream the laboratory scale BFB gasifier at ECN. Tar dewpoints
between 25°C and 170°C could successfully be measured. After finishing the tests a pre design
for a commercial analyser was made. Finally, the market for the TDA was identified and
segmented in R&D groups, indirect co-combustion and stand-alone biomass gasification
installations.


Keywords
Tar, dewpoint analyser, measurement method, biomass gasification.




2                                                                              ECN-C--05-026
CONTENTS

LIST OF TABLES                                                 4
LIST OF FIGURES                                                4
SUMMARY                                                        5
SAMENVATTING                                                   7
1.    INTRODUCTION                                             9
      1.1  Background                                          9
      1.2  Problem definition                                  9
      1.3  Objective                                          10
      1.4  Approach                                           11
2.    APPLICATION AND DESIGN                                  12
      2.1  Biomass gasification process                       12
      2.2  Specifications                                     13
      2.3  The dewpoint sensor                                15
      2.4  Gas conditioning section                           16
3.    VALIDATION TESTS                                        18
      3.1  Gas conditioning section                           18
           3.1.1 Tar condenser                                18
           3.1.2 Water removal section                        19
      3.2  Sensor fouling and recovery                        21
4.    TAR DEWPOINT ANALYSER PERFORMANCE                       23
      4.1  Measurement procedure                              23
      4.2  Operational test                                   23
      4.3  Influence of water on measurement                  26
      4.4  Gas conditioning section                           27
      4.5  Pre design of a commercial tar dewpoint analyser   27
5.    MARKET AND COMPETING TECHNOLOGIES                       28
6.    CONCLUSION                                              30
REFERENCES                                                    32




ECN-C--05-026                                                  3
LIST OF TABLES
Table 2.1    Representative product gas composition for an air blown fluidised bed
             biomass gasifier fuelled with demolition wood at a temperature of 850°C............13
Table 4.1    Comparison between the measured and calculated tar dewpoints. .......................24




LIST OF FIGURES
Figure 1.1   General layout of the tar dewpoint analyser (TDA)..............................................10
Figure 2.1   Layout of the ECN biomass gasification process..................................................12
Figure 2.2   Hydrocarbon and water dewpoints in a biomass gasification process
             compared with natural gas. .................................................................................14
Figure 2.3   The tar dewpoint sensor in the oven.....................................................................15
Figure 2.4   Layout of the gas conditioning section.................................................................16
Figure 3.1   Comparison of the calculated tar dewpoint at the outlet of the tar condenser
             and the temperature of the tar condenser. The calculations were performed
             with the ECN dewpoint model..............................................................................19
Figure 3.2   Performance of the water removal section at different operating
             temperatures. The operating temperature was similar to the water dewpoint.
             The reduction in water dewpoint corresponds to the difference in the
             temperature of the vaporisation section and the water dewpoint at the outlet.......20
Figure 3.3   Amount of tar deposits on the bed of beads at the 100°C and 280°C test run........22
Figure 4.1   Example of a tar dewpoint curve. ........................................................................24
Figure 4.2   Comparison of the calculated tar dewpoint with the measured tar dewpoint.
             On the Y=X line the calculated tar dewpoints equal the measured tar
             dewpoints............................................................................................................25
Figure 4.3   Tar dewpoint curves............................................................................................26
Figure 4.4   Pre design of the gas conditioning section and TDA for tar dewpoint
             measurements between 25°C and 170°C..............................................................27




4                                                                                                            ECN-C--05-026
SUMMARY
Biomass gasification is an attractive solution for the production of electricity and liquid or
gaseous transportation fuels from biomass. Currently, the short-term application of the product
gas is co-combustion in a coal power plant. The product gas replaces a part of the pulverised
coal. A more advanced application, for the mid-term, is the use of the product gas as fuel for gas
engines and gas turbines. For these applications the product gas needs to be cooled down and
de-dusted and the condensable tars must be removed. An advanced long-term application of the
product gas is the production of liquid and gaseous transportation fuels. For these advanced
applications the gas should be cooled down, de-dusted and tar should extensively be removed.

It is widely recognised that tars, a common name for light as well as heavy (aromatic)
hydrocarbons, are the most problematic species in product gas derived from biomass
gasification. Tar easily can lead to fouling due to condensation, when the product gas
temperature drops below the tar dewpoint. At present, counter measures for the prevention of tar
related fouling in a gasification process can only be taken when fouling has already occurred. In
practice, fouling is detected upon visual inspection of piping, during maintenance or when
severe fouling has resulted in blockage of piping or malfunctioning of equipment. An on-line
measurement method in biomass product gases for the detection of tar condensation is not
available as yet.

The present project is aimed at the development of an analyser for the on-line measurement of
the tar dewpoint1. Such an analyser can be implemented in a tar control system for the
prevention of fouling in biomass gasification processes. Three relevant operating conditions are
defined for the analyser. For the application in raw product gas, case 1, the existing sensor for
hydrocarbon dewpoint measurements in natural gas must be modified for the operation at high
temperature (60°C-350°C). Downstream a water-based tar removal system, case 2, the tar
dewpoint is close to the water dewpoint (20°C-60°C). Therefore water must be removed in the
gas conditioning section to prevent interference with the tar dewpoint measurement system.
Downstream advanced tar removal methods, case 3, the tar dewpoint is far below the water
dewpoint (<60°C) and water should be extensively removed in the gas conditioning section.

For the application of the existing hydrocarbon dewpoint analyser in biomass product gases the
hardware of the sensor was modified and a gas conditioning section was designed. The original
fibre-optics were replaced by high temperature resistant (250°C) fibre optics. The synthetic
material in the original sensor was replaced by ceramics. The internal diameter of piping was
increased to prevent blockage of the sensor by tar and the sensor cell was placed in an oven to
avoid cold spots in the sensor. The gas conditioning section consists of a filter, tar condenser
and water removal section. The filter removes dust and alkali metals in the system. The tar
condenser is at normal operation inactive but protects the water removal section or the dewpoint
sensor against heavy tar deposits. The gas drying section is installed to remove water from the
product gas when the tar dewpoint is close to or lower than the water dewpoint (cases 2 and 3).
A maintenance period of 3 months is defined as a minimum period for the gas conditioning
section.

The operational proof of principle test with the dewpoint analyser (sensor and gas conditioning
section) was successful. The tar dewpoint sensor responds strongly to tar condensate and the gas

1
    The tar dewpoint is the temperature at which the first tar molecules can condense out. In other words the tar
    dewpoint is the temperature at which the real total partial pressure of tar equals the saturation pressure. That does
    not mean that condensation will instantly occur, as the kinetics of this process can be slow, leading to over
    saturation. The dewpoint is dependent on the molecular mass of the compound and secondly on the concentration
    of the compounds.


ECN-C--05-026                                                                                                          5
conditioning section protect the dewpoint sensor against fouling and corrosion. The shape of the
dewpoint curve is similar to the shape of hydrocarbon (HC) dewpoint curves in natural gas. This
means that the HC dewpoint analyser is also applicable in biomass product gases. The tar
dewpoint analyser has measured tar dewpoints at tar concentrations between 2 and 12 g/mn3.

The tar dewpoint analyser accurately measured tar dewpoints in a temperature window from
25°C to 170°C. The maximum temperature of operation is limited by the thermal resistance of
the fibre optics to a tar dewpoint of 200°C. The current tar dewpoint analyser was successfully
tested for the application downstream tar removal technologies, case 2 and 3. For this
temperature window the instrument is ready for the next phase of development, the proof of
concept.

The proof of concept tests should be performed with a pre commercial design of the dewpoint
analyser measuring in clean product gas downstream a tar removal unit. Therefore, the
measurement and gas conditioning section should be automated. An automated tar dewpoint
analyser will sample and measure continuously and will display the tar dewpoint automatically
after a measurement cycle is finished. For a continuous operation of the gas conditioning
section, the static filter should be replaced by a regenerative filter. The tar condenser should
automatically dispose tar deposits and/or water condensate. Finally the pump downstream the
gas conditioning section should be replaced with a pump that can be operated above the tar
dewpoint. For high temperature measurements (> 200°C) the fibre optics should be cooled or
replaced by high temperature resistant fibre optics.

The market for the tar dewpoint analyser in biomass gasification processes was identified and
can be split into three segments:
1. R&D groups
2. Stand–alone biomass gasification systems
3. Indirect co-combustion
The markets can be separated in a short term and long-term market. R&D groups and Co-
combustion plants are potential end-users of the tar dewpoint analyser for the short-term market.
The market for the application in stand alone gasification systems is expected to expand in time
and is therefore identified as a long-term market. The current tar dewpoint instrument with a
temperature window between 25°C and 200°C is applicable in the R&D and stand-alone
markets. For the co-combustion market the sensor must be modified for the operation at high
temperature.




6                                                                               ECN-C--05-026
SAMENVATTING
Via de thermische vergassingsroute kan uit biomassa diverse energiedragers worden
geproduceerd zoals elektriciteit, vloeibare en gasvormige transport brandstoffen. Op dit moment
wordt het productgas afkomstig van een biomassa vergasser voornamelijk bij gestookt op een
ketel ter vervanging van de originele brandstof. In een poederkool centrale vervangt dit gas een
deel van de poederkool voor de productie van elektriciteit. Voor de middellange termijn is de
verwachting dat het productgas wordt toegepast in een gasmotor of gasturbine. Op de lange
termijn wordt naar verwachting biomassa via de vergassingsroute ook omgezet in vloeibare en
gasvormige transportbrandstoffen. Voor de middellange en lange termijn toepassingen dient het
productgas te worden afgekoeld, ontstoft en dient teer uit het gas te worden verwijderd.

Teer wordt over het algemeen gezien als een problematische groep van aromatische
koolwaterstoffen in productgas na een biomassa vergasser. Teer kan leiden tot vervuiling van
apparaten en leidingen zodra de productgas temperatuur daalt tot onder het teer dauwpunt. Op
dit moment kunnen maatregelen tegen teer gerelateerde vervuiling alleen worden genomen
nadat de vervuiling al is opgetreden. In de praktijk wordt de vervuiling vaak pas geconstateerd
tijdens een visuele inspectie of tijdens het groot onderhoud van installaties of als de vervuiling
al heeft geresulteerd in het verstoppen van pijpleidingen of apparaten. Om teer gerelateerde
vervuiling van apparaten tijdig te kunnen voorkomen is een on-line meetmethode gewenst die
de condensatie temperatuur van teer kan meten. Zo’n meetmethode bestaat nog niet.

Het doel van het onderhavige project is de ontwikkeling van een meetmethode voor het on-line
meten van de teercondensatie temperatuur, ook wel het teer dauwpunt genoemd. Een dergelijk
meetinstrument kan deel uitmaken van een controle systeem tegen de vervuiling van
procesapparatuur met teer. Voor de ontwikkeling van het meetinstrument, TDA (Tar Dewpoint
Analyser), zijn drie relevante condities gedefinieerd. Voor de toepassing van de TDA in ruw
productgas, cases 1, dient een bestaande sensor te worden aangepast voor bedrijf bij hoge
temperatuur (60°C tot 350°C). Na een water gebaseerde teer verwijdering is het productgas
verzadigd met teer en water bij een gas temperatuur variërend tussen de 20°C en 60°C. Voor het
meten van een teerdauwpunt in dergelijk productgas, cases 2, dient een deel van het water uit
het gas te worden verwijderd om verstoring van de dauwpuntsmeting met water te voorkomen.
Wanneer de TDA wordt toegepast na een intensieve teerverwijdering met een oliewasser of
katalytische teerkraker, cases 3, dient water uitgebreid te worden verwijderd omdat het teer
dauwpunt dan vaak veel lager is dan het water dauwpunt.

Voor de teer dauwpuntsmetingen is de hardware van een bestaande koolwaterstoffen
dauwpuntsmeter 2 aangepast en een gas conditionering ontworpen. De originele glasvezel kabels
van de dauwpuntsmeter zijn vervangen door kabels die bestand zijn tegen een hoge temperatuur
(250°C). Verder is al het synthetische materiaal vervangen door keramiek. Bovendien zijn de
diameters van de leidingen vergroot om verstopping van de leidingen te voorkomen. Om koude
plekken in de sensor te voorkomen is de gehele sensor in een oven geplaatst. De gas
conditionering is ontworpen om de sensor te beschermen tegen vervuiling en corrosie. De gas
conditionering bestaat uit een filter, teer condensor en water verwijderingsectie. In het filter
worden deeltjes en alkali zouten verwijderd. De teer condensor is tijdens normaal bedrijf niet
actief maar beschermd de water verwijderingsectie en de sensor tegen vervuiling met zwaar teer
die kan ontstaan tijdens onstabiel bedrijf van het biomassa vergassingsproces. De water
verwijderingsectie wordt gebruikt op het moment dat het teer dauwpunt lager of gelijk is aan het
water dauwpunt (cases 2 en 3). Verder is voor de gas conditionering een onderhoudscyclus van
minimaal 3 maanden gedefinieerd.

2
    De bestaande dauwpuntsmeter wordt ingezet voor het meten van koolwaterstof dauwpunten in aardgas.


ECN-C--05-026                                                                                           7
De operationele “proof of principle” test met de TDA was succesvol. De sensor reageert sterk
op teer condensaat en de gas conditionering beschermd de sensor tegen vervuiling en corrosie.
De vorm van de dauwpuntscurve is identiek aan de koolwaterstoffen dauwpuntscurve in
aardgas. Dat betekent dat de dauwpuntsmeter geschikt is voor het meten van teer dauwpunten.
De TDA heeft teer dauwpunten gemeten bij teerconcentraties variërend tussen de 2 en 12 g/mn3.

De huidige TDA is in staat om teerdauwpunten in een temperatuurbereik tussen de 25°C en
170°C nauwkeurig te meten. Het maximale meetbereik van 200°C wordt momenteel bepaald
door de thermische weerstand van de fibre optics. De huidige TDA is succesvol getest voor de
toepassing na teer verwijderingsapparaten, cases 2 en 3. Voor deze toepassing is de TDA klaar
om de volgende “proof of concept” fase in te gaan.

De “proof of concept” dient uitgevoerd te worden met een commercieel voorontwerp voor
metingen in schoon product gas na teer verwijderingsapparaten. In deze fase dient zowel de
metingen als de gas conditionering van de TDA volledig te zijn geautomatiseerd. Een
geautomatiseerde TDA meet continu en geeft het teerdauwpunt automatisch weer op een
scherm. Voor een continu bedrijf van de gas conditionering dient het statische filter te worden
vervangen door een regeneratief filter. Ook dient de pomp na de TDA te worden vervangen met
een pomp die boven het teer dauwpunt kan worden bedreven.

De markt voor de TDA kan worden opgedeeld in drie segmenten:
1. R&D groepen
2. Stand-alone biomassa vergassingssystemen
3. (Bij)stook van het productgas op ketels
De markt segmenten kunnen worden opgedeeld in een korte termijn en lange termijn markt. De
R&D groepen en (bij)stook toepassing zijn potentiële eindgebruikers van de TDA voor de korte
termijn markt. De markt voor stand-alone biomassa vergassingssystemen neemt naar
verwachting in omvang toe en wordt daarom gezien als een lange termijn markt. De huidige
TDA met een temperatuurbereik tussen de 25°C en 200°C kan worden toegepast in de R&D
groepen en stand-alone biomassa vergassingsmarkt. Voor toepassing in de (bij)stook markt
dient de sensor geschikt te worden gemaakt voor bedrijf bij hoge temperatuur.




8                                                                             ECN-C--05-026
1.         INTRODUCTION
1.1        Background
Biomass gasification is an attractive technology for the production of electricity and liquid or
gaseous transportation fuels from biomass. Currently, the short-term application of the product
gas is indirect co-combustion in coal power plants. The product gas replaces a part of the
pulverised coal after intermediate cooling and de-dusting. A more advanced application, for the
mid-term, is the use of the product gas as fuel for gas engines. For this application the product
gas needs to be cooled down and de-dusted and the condensable tars must be removed. An
advanced long-term application of the product gas is the production of liquid and gaseous
transportation fuels. For these advanced applications the gas should be cooled down, de-dusted
and tar should extensively be removed.

It is widely recognised that tar3 is the most problematic specie in product gas derived from
biomass gasification [1]. In practice the product gas from a gasifier will contain 1-20 g/mn3 of
tar, depending on the type of gasifier and conditions applied. Due to the wide spectrum of tar
compounds present in the product gas, they may condense over a wide temperature range
normally starting at a temperature above 200°C. Condensation of tar may initiate fouling in
equipment and can ultimately result in plugging of piping, coolers [2], filters, and (packed)
towers.

The first step in the prevention of fouling is the development of tar removal systems. Promising
tar removal methods under development are catalytic tar reformers [3], oil scrubbers [4] and
water scrubbers [5]. The latter technology results in a product gas that is saturated with tar,
which can be applied as fuel for a gas engine [6], but has the disadvantage of the production of
tar contaminated wastewater. For the production of liquid and gaseous transportation fuels, tar
must be removed (almost) completely by oil scrubbing or in a catalytic reformer, to reduce the
tar dewpoint, i.e. the temperature at which tar starts to condense, to below -5°C.

The next step in the prevention of fouling is the development of a tar control system. It is
expected that a timely measurement of tar condensation is crucial in the prevention of tar related
fouling. When condensation is detected, proper counter measures can be taken to prevent
damage to or shutdown of the installation. The heart of such an integrated “tar control system”
is the tar condensation sensor. To date, no suitable sensor is available for the measurement of tar
condensation.


1.2        Problem definition
At the moment, counter measures for tar related fouling in a gasification process can only be
taken when fouling has already occurred. In practice, fouling is detected upon visual inspection
of piping during maintenance or when severe fouling has resulted in blockage of piping or
malfunctioning of equipment. An on-line measurement method for the determination of the tar
condensation temperature in biomass product gases is not available yet.




3
    Tar is a common name (used here) for aromatic hydrocarbons with a molecular mass larger than benzene.


ECN-C--05-026                                                                                               9
1.3    Objective

The present project is aimed at the development of a tar dewpoint analyser (TDA) for the on-
line measurement of tar dewpoints.

The tar dewpoint is the highest temperature, at which the first tar molecules can condense out.
In other words the tar dewpoint is the temperature, at which the real total partial pressure of tar
equals the saturation pressure. That does not mean that condensation will instantly occur, as the
kinetics of this process can be slow, leading to an over saturation of the product gas with tar.
The dewpoint is dependent on the molecular mass of the compound and secondly on the
concentration of the compounds.

The on-line TDA consists of a gas conditioning section and a sensor. Figure 1.1 shows the
general layout of the TDA. The scope of the project is the design of a dedicated gas
conditioning section to protect the tar dewpoint sensor against fouling and corrosion, and the
modification of an existing dewpoint sensor for tar dewpoint measurements.

                                                              TAR
                                                            DEWPOINT

                                                             DISPLAY



            Biomass gasification      Gas                                         Vent gas or
     MAIN                                                   Dewpoint
                                   conditioning
     DUCT   product gas sampling                             Sensor        recycle to the main duct
                                     section


                               ECN SCOPE               MICHELL INSTRUMENT SCOPE



Figure 1.1 General layout of the tar dewpoint analyser (TDA).

The TDA can be implemented in a tar control system for the prevention of tar related fouling in
biomass gasification processes. Three commercially relevant applications were defined for the
TDA:
1. A biomass gasification process without a tar removal system for the co-combustion of the
   product gas in a coal boiler. An example of such a process is the 80 MWth biomass
   gasification plant from Essent at the AMER, Geertruidenberg in Holland. In this process the
   TDA can be used in the protection of the primary gas cooler against tar related fouling.
   Temperature window for tar dewpoint measurements is between 60°C and 350°C.
2. A biomass gasification process with an aqueous scrubbing system for tar removal and a gas
   engine for the production of electricity. This configuration is used e.g. in Wiener Neustadt
   and Harboore with fixed bed gasifiers. The tar dewpoint analyser can be used in the
   protection of the gas engine against tar related fouling. The temperature window for tar
   dewpoint measurements in this application is between 20°C and 60°C.
3. A biomass gasification process with an oil scrubbing system or a catalytic reformer for tar
   removal and a gas engine for the production of electricity, like the biomass gasification
   plant in Guessing, Austria. The tar dewpoint analyser can be used in the protection of the
   gas engine against tar related fouling. Temperature window for tar dewpoint measurements
   is between –15°C and 60°C.




10                                                                                  ECN-C--05-026
1.4    Approach

ECN and Michell Instruments are partners in the project. Michell Instruments is the leading firm
in the water dewpoint and gas sensor market. Starting point in the project for Michell
Instruments is the modification of an existing dewpoint sensor, which is a patented technology
for hydrocarbon dewpoint measurements in natural gas, to make it suitable for tar dewpoint
measurements in product gas from a biomass gasifier. ECN has to design a dedicated gas
conditioning system to prevent fouling and corrosion of the sensor.

The project was divided in 7 work packages:
1. Project management
2. Draw up of specifications
3. Design and construction
4. Laboratory tests
5. Pre design of a commercial tar analyser
6. Marketing plan
7. Reporting

The work has been executed according to the project proposal. On basis of the specifications,
operating temperature window and gas characteristics, an existing hydrocarbon dewpoint sensor
was modified and a dedicated gas conditioning section was designed. Preliminary laboratory
tests with the gas conditioning section and dewpoint sensor were run to investigate the
performance and fouling characteristics of the dewpoint sensor and the gas conditioning section.
The tar dewpoint analyser (= sensor + gas conditioning section) was tested downstream the
laboratory scale gasifier at ECN. After the laboratory tests a pre design for a commercial
analyser was made. The tar dewpoint analyser was compared with other tar measurement
methods. Finally, the market for the tar dewpoint analyser was identified and segmented,
according to (possible) applications.




ECN-C--05-026                                                                                11
2.         APPLICATION AND DESIGN
In natural gas the dewpoint of hydrocarbons (HC) (up to C10 compounds) is determined to
protect distribution pipelines against fouling due to the condensation of hydrocarbons. Michell
Instruments is the only supplier of hydrocarbon dewpoint meters. This project aims at the
development of a HC dewpoint instrument for the application in biomass product gases, to
protect equipment and pipelines against fouling due to the condensation of tar. For the
measurement of tar dewpoints in biomass product gases the original HC dewpoint sensor must
be modified and a gas conditioning section must be installed upstream the analyser.

This chapter will discuss the design of the tar dewpoint sensor, the gas conditioning section and
the biomass gasification process, in which the TDA will be applied.


2.1        Biomass gasification process
A general layout of the biomass gasification process is given in Figure 2.1. The scheme
represents the ECN gasifier and gas cleaning system. Inside the Circulating Fluidised Bed
(CFB) gasifier biomass is converted with air into a product gas at approximately 850°C. The
product gas is mainly composed of CO, H2, CH4, CO2, N2 and H2O. However, beside the main
products, by products like tar4, NH3 and dust are also present. A typical composition of the
product gas is given in Table 2.1.




Figure 2.1 Layout of the ECN biomass gasification process.

Dust is removed from the product gas downstream the cooler. The de-dusted gas is fed to the
OLGA for the removal of tar with washing oil. Downstream the OLGA, NH3 is removed from
the product gas in an aqueous scrubber. After NH3 removal the product gas is clean enough for
the application in a gas engine. When the gas is used as fuel for a boiler, tars and NH 3 are
normally not removed.




4
    Tar is a general name of mainly aromatic hydrocarbons with a molecular mass larger than benzene.


12                                                                                               ECN-C--05-026
Table 2.1 Representative product gas composition for an air blown fluidised bed biomass
          gasifier fuelled with demolition wood at a temperature of 850°C.
                                                                   wet basis
 CO                                 vol%                               16
 H2                                 vol%                               14
 CO2                                vol%                               14
 CH4                                vol%                                4
 N2                                 vol%                               36
 H2O                                vol%                               13
 C2H4                               vol%                              1.4
 C2H6                               vol%                              0.01
 Benzene                            vol%                              0.4
 Toluene                            vol%                              0.1
 H2S                                ppmv                           100-200
 COS                                ppmv                                3
 NH3                                ppmv                          2000-3000
 HCl                                ppmv                            75-150
 HF                                 ppmv                             10-20
 Tar                               mg/mn3                            7000
 Ash                                g/mn3                              25

In the biomass gasification process the dewpoint analyser can be applied to protect equipment
and piping against fouling due to the condensation of tar. Tar condensation in the equipment
depends on the actual tar dewpoint and the (local) temperature of the equipment. When the
temperature of the equipment is below the tar dewpoint, the gas is over saturated with tar and tar
is prone to condense, provided sufficiently long residence time and contact surface are
available.

Fouling of equipment due to the condensation of tar has shown to be a problem in biomass
gasification processes. Potential locations with a high risk for tar condensation are the cold
surfaces of the hot gas cooler, the NH3 scrubber and the gas engine. The TDA can help in
protecting the equipment of the installation against tar deposition. The analyser can measure the
tar dewpoint and compare the dewpoint with the operating temperature. As soon as the tar
dewpoint approaches the operating temperature of the equipment, the analyser should give an
alarm, so that counter measures can be taken to prevent tar deposition inside the equipment. The
analyser can also form a part of a process control system.

The tar concentration and composition in a biomass gasification process are strongly dependent
on the type of gasifier and the tar removal technology used. Tar from a fixed bed updraft
gasifier is normally more reactive, i.e. easily converted, but it also condenses at a lower
temperature than tar from a fluidised bed gasifier. Nevertheless, the tar concentration in product
gas downstream an updraft gasifier can be 10 times higher than in product gas downstream a
fluidised bed gasifier [7]. The tar concentration downstream OLGA or a catalytic tar cracker can
be 5 times lower than the tar concentration downstream a water based tar removal system. The
differences in tar compositions and concentrations implicates that the TDA should be able to
measure at high as well as at low tar concentrations and at a wide temperature range between –
20°C and 350°C.


2.2    Specifications
The characteristics of product gas from a biomass gasifier differ significantly from natural gas.
In natural gas the HC dewpoint varies between -20°C and 15°C and lies above the water
dewpoint and the pressure varies between a few and hundred bars. The pressure of the product
gas is usually lower, in the range of 1-20 bars. Even more significant is the temperature window


ECN-C--05-026                                                                                  13
for the tar dewpoint in product gas. In raw product gas the tar dewpoint can be as high as 350°C.
After tar removal the tar dewpoint can decrease to -20°C to 60°C, depending on the tar removal
system applied. This means that the tar dewpoint window for the dewpoint analyser must be
between –20°C and 350°C, which is significantly higher than the operating temperature in
natural gas.

The gas composition is another important difference between natural gas and biomass product
gas. Biomass product gas is mainly composed of H2, CO, CO2, N2, CH4 and water.
Condensation of water starts at a temperature between 50°C and 65°C. The product gas can also
contain by products like dust, alkali metals, NH3, H2S, COS and HCl, which are usually absent
in natural gas. Therefore, a gas conditioning section must be applied to protect the sensor
against fouling and corrosion and to avoid water condensation in the sensor.

Three commercially relevant operating conditions were specified that lead to the following
requirements:
1. For the application in raw product gas, case 1 in the objective, the sensor must be modified
    for operation at a high temperature (60°C-350°C).
2. Downstream a water-based tar removal system, case 2 in the objective, the tar dewpoint lies
    between 20 and 60°C and is comparable to or slightly higher than the water dewpoint.
    Therefore water must be removed in the gas conditioning section to prevent interference of
    the measurement by water.
3. Downstream advanced tar removal methods, case 3 in the objective, the tar dewpoint lies
    between –15°C and 60°C, but always below the water dewpoint. Therefore, water should
    extensively be removed in the gas conditioning section.
Figure 2.2 summarises graphically the difference in conditions between the three applications
mentioned and natural gas.


                                                            400
                                                                                     case 1
                                                            300
     Hydrocarbon dew point [°C]




                                                            200


                                        natural gas         100
                                                                                   case 2          case 3
                                                              0
                                  -60       -40       -20          0     20         40        60        80
                                                            -100


                                                            -200
                                                            Water dew point [°C]

Figure 2.2 Hydrocarbon and water dewpoints in a biomass gasification process compared with
           natural gas.

The time necessary for a tar dewpoint measurement defines the response delay for counter
measures. Tar induced fouling can be minimised when the measurement time is short, so
counter actions, to decrease the tar dewpoint, can be taken in time. Therefore, a short duration of
a measurement cycle is a prerequisite in the prevention of tar induced fouling.


14                                                                                                          ECN-C--05-026
2.3     The dewpoint sensor
The sensor cell is the key element of the dewpoint analyser. A beam of light is focused at the
centre of an optical surface. The reflection of the light beam is measured. The difference
between the intensity of the original light beam and its reflection, here upon called the signal, is
used for the detection of the HC dewpoint. As soon as the signal reaches a certain threshold
value, the corresponding surface temperature of the optical surface is displayed as the dewpoint
of the hydrocarbons in the gas. This threshold value is determined practically.

A HC dewpoint measurement cycle in natural gas starts with flushing the sensor cell with
natural gas. When the cell is filled with fresh natural gas the gas flow is stopped by means of
valves, thus closing off the cell. The optical surface of the sensor is than slowly cooled down,
using compressed air or compressed natural gas. As soon as hydrocarbons condense on the
optical surface, the reflection of the light beam is changed by the hydrocarbon condensate. The
optical surface is recovered by increasing the temperature and flushing the sensor with natural
gas. After recovery the next measurement cycle can start. Such a measurement cycle takes a few
minutes.

The hardware of the HC dewpoint analyser was modified for tar dewpoint measurements. The
original fibre-optics was replaced by high temperature resistant (250°C) fibre optics. The
synthetic material in the original sensor was replaced by ceramics. The internal diameter of
piping was increased to prevent blockage by tar. The sensor cell was placed in a stove to avoid
cold spots in the sensor. The modified sensor is shown in Figure 2.3.




Figure 2.3 The tar dewpoint sensor in the oven.

Also the measurement procedure was changed for the tar dewpoint measurements. These were
manually performed with a continuous product gas flow through the cell. The measurement
procedure is further described in chapter 4.




ECN-C--05-026                                                                                    15
2.4     Gas conditioning section
The gas conditioning section is installed upstream the measurement cell to prevent the sensor
from fouling and corrosion with dust, alkali metals and soot/heavy tar. The layout of the gas
conditioning section is schematically given in Figure 2.4. The gas conditioning section is
composed of a filter, tar condenser and water removal section (dryer). The filter removes dust
and alkali metals in the system. The tar condenser is a guard filter, which protects both the water
removal section and the dewpoint sensor against heavy tar deposits that should not occur in
normal operating conditions. The dryer removes water from the product gas when the tar
dewpoint is equal or lower than the water dewpoint (case 2 and 3). A maintenance period of 3
months is defined as a minimum period for the gas conditioning section.


                                                    If tar        Water removal
                                                dewpoint<80°C
                                                                                         Gas to tar
                                                                                         dewpoint     f
                                                                                          sensor


               Dust removal              Tar condenser
                                                                         If tar
                                                                     dewpoint>80°C



Figure 2.4       Layout of the gas conditioning section.

As described above the product gas is first drawn through a dust filter, to prevent fouling of the
sensor cell with dust and alkali metals. Even for measurements in de-dusted product gas a filter
will be installed to protect the sensor from the fine particles that might be present in the gas. The
alkali metal aerosols are removed together with the dust in the filter. Since a static filter has a
limited stand time, which is further dependent on the dust load in the product gas, a regenerative
filter is foreseen to increase maintenance intervals. For the proof of principle tests, a static filter
was still used.

Downstream the dust filter the product gas is passed through the tar condenser. At steady state
the tar dewpoint in the product gas is lower than the operating temperature of the tar condenser.
Thus at normal operation tar does not condense in the tar condenser. Due to disturbances in the
gasification process, heavy tar with an excessively high dewpoint can be present in the product
gas, and the TDA gives an alarm to warn for the increasing risk of tar induced fouling in the
biomass gasification process. Subsequently, the tar condenser captures the heavy tars to protect
the downstream water removal section or the sensor against tar deposits that can not be re-
vaporised. Thus the motive for the installation of a tar condenser is dependent on the
application:
• For measurements in raw product gas, the tar condenser is installed upstream the sensor to
    protect the sensor against fouling with heavy tars. Heavy tar deposits on the optical surface
    need a high temperature of recovery. The thermal resistance of the materials of the sensor
    cell determines the maximum temperature of recovery. Thus in case 1 the condenser
    protects the sensor against irreversible fouling of the optical surface with heavy tars that can
    not be vaporised during the recovery phase, due to temperature limitations of the materials.
• For measurements downstream a tar removal system, the tar condenser must protect the
    water removal section against the condensation of tars. The tar dewpoint in product gas
    downstream tar removal units is normally between 0°C and 60°C. To prevent fouling of the
    water removal unit with tar, the condenser removes all the tar that condenses above the
    operating temperature of the water removal section, which operating temperature is 30°C
    higher than the actual tar dewpoint in the gas.




16                                                                                   ECN-C--05-026
The water removal section is installed when the tar dewpoint in the product gas is lower or
equal to the water dewpoint, thus typically in product gases downstream tar removal units (cases
2 and 3). When water condenses on the sensor, the tar dewpoint measurement will be disturbed.
Therefore, the water dewpoint should be decreased in the water removal section to well below
the tar dewpoint. To avoid tar condensation in the water removal section, it is operated at a
temperature, which is higher than the actual tar dewpoint.




ECN-C--05-026                                                                                17
3.      VALIDATION TESTS
Preliminary tests were performed to validate the gas conditioning section and to investigate
potential fouling issues in the gas conditioning section and the sensor of the TDA.


3.1     Gas conditioning section
In the design of the gas conditioning section the tar condenser and water removal section were
identified as potential bottlenecks. Therefore, preliminary tests were performed with the water
removal section and the tar condenser. The potential sources of problems for the gas
conditioning section are:
1. The formation of tar aerosols in the tar condenser
2. Fouling of the water removal section due to the condensation with tar
The preliminary test results with the tar condenser and the water removal unit in the gas
conditioning section are reported in the following section.


3.1.1 Tar condenser
The tar condenser must protect the water removal section and the sensor cell against tar related
fouling. The tar condenser will only remove increased concentrations of heavy tars, which can
be formed due to disturbances in the biomass gasification process. These heavy tars easily form
tar aerosols, which can only be captured with an absolute (aerosol) filter. However, the
installation of a filter for the capture of tar aerosols is undesirable, because an absolute filter is
easily blocked with tar aerosols and thus would lead to a significant increase in the maintenance
frequency for the gas conditioning section. Therefore, the tar condenser must remove heavy tars
and avoid the formation of tar aerosols.

The tar condenser was tested with de-dusted product gas from the ECN 1 kg/h BFB gasifier.
The tar condenser was operated at several temperatures. At each temperature, the tar
concentration downstream the condenser was measured using the SPA method [8]. The tar
dewpoint was calculated with the ECN model using the concentrations of individual tar
compounds. Comparison of the tar dewpoint with the temperature of the condenser was used for
the determination of the presence of tar aerosols in the product gas. When aerosols are present,
the tar dewpoint will be higher than the temperature of the tar condenser. The condenser is
performing well when the tar dewpoint at its outlet is equal to the operating temperature of the
condenser.




18                                                                                  ECN-C--05-026
                              200



        Tar dewpoint [°C] s
                              160




                              120




                               80
                                    80      120               160              200
                                         Temperature tar condenser [°C]

                                          Y=X line       Dew point [°C]
Figure 3.1 Comparison of the calculated tar dewpoint at the outlet of the tar condenser and the
           temperature of the tar condenser. The calculations were performed with the ECN
           dewpoint model.

Figure 3.1 shows the results of the tar condenser test. An ideal performance of the tar condenser
is given by the Y=X line. The dewpoints on this line are equal to the temperature of the
condenser. From Figure 3.1 it is clear that the tar dewpoints are all arranged approximately 3°C
of the Y=X line. This means that the tar condenser performs well and tar aerosols are removed
or not formed in the condenser.


3.1.2 Water removal section
The water removal section is installed to avoid disturbance of the tar dewpoint measurements by
water. This section is used in applications case 2 and 3 for tar dewpoint measurements between
–20°C and 60°C, i.e. temperatures equal to or below the actual water dewpoint in the product
gas. The water removal section must reduce the water dewpoint below the tar dewpoint. Both
the performance and potential risk of tar related fouling were investigated with synthetic gases.

Performance
The performance of the water removal section was investigated at several operating
temperatures and water concentrations at the inlet. In a vaporisation unit (using impingers) N2
was saturated with water. The temperature of the vaporisation unit determined the water
concentration and dewpoint at the inlet of the water removal section. The water removal section
was operated at the same temperature as the vaporisation unit. The experimental results are
given in Figure 3.2.




ECN-C--05-026                                                                                 19
                                 50

                                 40

                                 30
              Temperature [°C]
                                 20

                                 10

                                  0

                                 -10

                                 -20
                                       20                 40                  60                    80
                                                          Water dewpoint IN [°C]

                                            Water dewpoint OUT [°C]   Reduction of water dewpoint


Figure 3.2 Performance of the water removal section at different operating temperatures. The
           operating temperature was similar to the water dewpoint. The reduction in water
           dewpoint corresponds to the difference in the temperature of the vaporisation
           section and the water dewpoint at the outlet.

Figure 3.2 clearly illustrates the reduction of the water dewpoint in the water removal section.
The water dewpoint at the entrance was varied between 30°C and 70°C. The reduction in the
water dewpoint was constant at 40°C, and independent of the operating temperature. Therefore,
the dewpoint in the outlet increased from –10°C to 30°C when the temperature of the water
removal section increased from 30 to 70°C. The results are according to the specifications of the
supplier.

For the measurement of tar dewpoints below 20°C an extra measure should be taken in the gas
conditioning section for the water removal. The water dewpoint in product gas normally varies
between 50 and 60°C. Since, water may not condense in the water removal section, the
temperature of this section must stay above the water dewpoint. At an operating temperature of
60°C, the water removal section can reduce the water dewpoint to 20°C, which determines the
lower limit for tar dewpoint. To measure a tar dewpoint below a temperature of 20°C, a further
decrease in the water dewpoint is necessary. This further decrease can be accomplished by the
partial condensation of water in the tar condenser, which allows for a lower operating
temperature in the water removal section, which reduces the water dewpoint below 20°C.

Fouling
Tar-loaded gas downstream the tar condenser enters the water removal section. Since the water
removal section can decrease the water dewpoint with an average of 40°C, it should be operated
at a temperature that is close to the actual tar dewpoint. Therefore, cold spots inside the water
removal section are a potential risk for fouling. Therefore, in a preliminary test, fouling of the
water removal section was investigated, using phenol and naphthalene as representatives for
polar and non-polar tar compounds, respectively.

For the fouling tests phenol and naphthalene as well as water were vaporised in nitrogen gas.
The N2 was fed to the water removal section. The phenol and naphthalene concentrations were
measured at the inlet and the outlet of the water removal section. The water removal section was



20                                                                                                  ECN-C--05-026
operated at a temperature of 70°C. In two tests two gaseous mixtures with different tar
concentrations were fed to the water removal unit:
• N2 gas saturated with naphthalene and phenol at 70°C
• N2 gas with representative naphthalene and phenol concentrations for biomass product gases
   from a biomass fluidized bed gasifier operated at 850°C. The dewpoint of naphthalene was
   20-30°C below the operation temperature of the water removal section.

In the first test, approximately 60% of the phenol and naphthalene were lost in the water
removal section. In the second test the phenol and naphthalene concentrations did not change
significantly. Apparently the temperature of the water removal section in the second test was
high enough to prevent the condensation of naphthalene and phenol. Therefore, as long as the
temperature of the water removal unit is well controlled and 20 to 30°C above the tar dewpoint,
condensation of tar can be avoided.


3.2     Sensor fouling and recovery
Fouling of the sensor can be caused by tar polymerisation, as such a polymer does not vaporise
during the recovery phase of the sensor. Fouling of the optical surface with polymerised tar is
irreversible, leading to its destruction. Such tar deposits on the optical surface can not be
removed by increasing the temperature or decreasing the pressure during the recovery phase.

To investigate sensor fouling, product gas, - saturated with tar at 100°C and 280°C -, was drawn
over a packed bed of beads heated to temperatures of 100°C or 280°C. The bed of beads was
made of the same material as the optical surface. In both test runs, product gas with a flow rate
of 1 ln/min was drawn for approximately 15 minutes over the bed, which was subsequently
recovered in another 15 minutes by increasing the temperature of the beads with 30°C and
flushing with 1 ln/min of N2. During the recovery phase the tar vaporises and is carried away by
nitrogen. For both test runs (100 and 280°C) this measurement cycle was repeated four times.
After both test runs, the amount of tar deposit on the beads was determined by washing the
beads with a fixed amount of DCM. The amount of tar in the DCM extract was analysed by
means of a GC method for tar.

Upstream the bed of beads, dust was removed with a filter operated at 350°C and the product
gas was saturated with tar by gas cooling in a tar condenser operated at 100°C or 280°C. The tar
condenser and filter were installed in a side stream of the main duct to the flair. The gasifier was
operated at standard conditions of 850°C and with beech as feedstock. The temperature of the
tar condenser determined the tar dewpoint in the product gas.

The beads operated at 100°C remained visually clean but the DCM extract contained
approximately 20 mg of tar (Figure 3.3), which is approximately 5% of the total amount of tar
that was drawn over the bed during the 60 minutes of operation. At 100°C the tar deposits
contained mainly compounds with a molecular weight between phenanthrene and pyrene. These
compounds were condensed on the bed of beads during operation. During the recovery phase
these compounds should have been vaporised completely, which did not happen. Apparently the
conditions during the recovery phase were not sufficient for the complete vaporisation of the
compounds.

The beads operated at 280°C had a light yellow colour and the DCM extract contained 0.7 mg
of tar (Figure 3.3), which is approximately 0,1% of the total amount of tar that was drawn over
the bed during the 60 minutes of operation. Although the concentration of condensed tar was
low, the deposits were not expected at 280°C. Normally the compounds in the tar deposits
condense below a temperature of 220°C, which is far below the operating temperature of the
bed. The yellow colour can be explained either by the condensation of heavy tars or fouling by
polymerised tars that did not vaporise during the recovery phase. In case of polymerised tars the


ECN-C--05-026                                                                                    21
deposition would be irreversible, which would mean that there is a maximum temperature for
the tar dewpoint measurements.

                      6000


                      5000


                      4000
     Amount [ µ g]




                      3000


                      2000


                      1000


                            0




                                                                                           en
                                                                    Be ora en




                                                                                             e
                                                                                   Py n




                                                                                           en
                                      en




                                                                                           en
                                            en




                                                                    B e ( e ) - en
                                                                               C een
                                                                                           en




                                                                                   -p n
                                                                                Pe en
                                                     en




                                                                                            n




                                                                               )-p en
                                n




                                                                                         en
                                                                                           e




                                                                                           e
                                                                                         ee
                            ee




                                                               Be h)-a ryle
                                                                                         e
                                                                                       he
                                    re

                                           re




                                                                                       he
                                                                                       re




                                                                                       re

                                                                                         e




                                                                          gh race
                                                 ce




                                                                                        le
                                                                                       le
                                                                          -fl nth




                                                                                     on
                                                                                     ac




                                                                                      yr
                                                                                     ys
                            ht




                                                                                     ry
                                                                                     ry
                                                                                    py
                                 uo

                                           th



                                                          nt




                                                                                    nt
                                                ra
                       ap




                                                                                    e



                                                                                  pe
                                                                                  hr




                                                                                  or
                                                                                  hr
                                       an




                                                                                  h
                                                      ra




                                                                                 ra
                                 Fl



                                                th




                                                                               a)




                                                                               nt
                                                                               nt




                                                                               C
                       en




                                                                              i)-
                                                                             uo
                                                     uo
                                      en

                                            An




                                                                            o(



                                                                 en -c d
                                                                            -a




                                                                             u

                                                                            o
                     Ac




                                                                          -fl
                                                 Fl
                                  Ph




                                                                        nz

                                                                        nz
                                                                        a)




                                                                       o(
                                                            D 123
                                                                      b)

                                                                       k)




                                                                       a
                                                               o(




                                                                    z(
                                                                    o(
                                                                    o(




                                                                   nz
                                                           nz




                                                                   (
                                                                 nz
                                                                 nz




                                                                no
                                                          Be




                                                             Be
                                                             Be




                                                              ib
                                                            de
                                                          In


                                                                    beads 100°C   beads 280°C

Figure 3.3 Amount of tar deposits on the bed of beads at the 100°C and 280°C test run.

In conclusion, at the high as well as at the low temperature the beads were not completely clean
after recovery. For the 100°C test run an increase in recovery temperature probably solves this
problem. Increasing the recovery temperature in the 280°C test run further, might not solve the
problem, when the irreversible fouling was caused by polymerised tars. For the tar dewpoint
experiments the recovery temperature was set to 50°C above the actual tar dewpoint
temperature.




22                                                                                              ECN-C--05-026
4.         TAR DEWPOINT ANALYSER PERFORMANCE
The TDA, designed and built in this project, was installed downstream the laboratory scale
gasifier WOB at ECN. In this chapter the experimental results of the proof of principle test with
the TDA and the pre design of a commercial tar TDA analyser will be discussed.


4.1        Measurement procedure
The TDA was tested downstream the 1 kg/h bubbling fluidised bed gasifier, WOB, at ECN.
Raw product gas downstream the cyclone was continuously drawn with a pump downstream the
dewpoint analyser through the gas conditioning section and the sensor cell. Dust particles were
removed by means of a dust filter and the tar condenser was used to supply the sensor cell with
a product gas of different tar dewpoints. The temperature of the tar condenser determined the tar
dewpoint supplied to the sensor cell. The water removal section was only used for the
measurement of tar dewpoints below 60°C, when the tar dewpoint is nearing the water
dewpoint. The temperature of the oven as well as of the piping to the sensor was kept constant
at the same temperature as the tar condenser.

A measurement cycle started with turning on cooling air. Cooling of the optical surface
continued until a signal of more than 2000 mV was obtained. The signal and temperature of the
optical surface were recorded manually. From the dewpoint curve (signal versus optical surface
temperature), the actual dewpoint was determined graphically. The surface was recovered by
turning off the airflow manually, and increasing the temperature of the surface with the heaters
in the cell body of the sensor. During cleaning the tar deposits vaporise and are carried away by
the product gas. After the signal returned to its baseline, the heaters were turned off and the air-
cooling was started for the next dewpoint measurement. During the measurement cycle product
gas was continuously drawn through the gas conditioning section and sensor.

At each measurement cycle a tar sample was taken at the entrance of the oven. Tar sampling
was performed according to the SPA method. With the SPA method the concentrations of
individual tar compounds with a molecular mass between Xylene (M=106) and Coronene
(M=300) can be measured. From the tar spectrum a tar dewpoint can be calculated with the
ECN dewpoint model. The accuracy of the calculation was validated in another project and is
approximately 3°C at a temperature window of 100 to 170°C. The model was not validated at a
lower temperature but is expected to become more accurate5. The calculated tar dewpoint will
be compared with the measured tar dewpoint to validate the tar dewpoint measurements.


4.2        Operational test
Figure 4.1 shows a typical dewpoint curve obtained during a tar dewpoint measurement. The
measurement starts at high temperature and a signal between 0 and 200mV. At first, the signal is
noisy but stable around a baseline. Approximately 10°C above the estimated dewpoint
temperature, the signal starts to increase gradually. Close to the dewpoint temperature the signal
increases exponentially. The measurement was stopped at a signal of 2000mV. The tar dewpoint
was estimated graphically from the dewpoint curve. The procedure is given in Figure 4.1. Two
lines of contact are drawn on the curve. The point of intersection of these straight lines
represents the average tar dewpoint.


5
    Due to the availability of a wide range of vapour pressure curves and the high concentrations of tar compounds, it
    is expected that the accuracy of the tar dewpoint model is improving below a temperature of 100°C.


ECN-C--05-026                                                                                                     23
                       2200

                       2000

                       1800

                       1600
  Signal Change [mV]




                       1400

                       1200

                       1000

                        800

                        600

                        400

                        200

                          0
                           50.0   55.0   60.0   65.0       70.0      75.0     80.0         85.0   90.0    95.0   100.0
                                                       Optical Surface Temperature, [°C]

Figure 4.1 Example of a tar dewpoint curve.

Table 4.1 and Figure 4.2 compare the calculated tar dewpoints with the measured ones. The
total tar concentration during the TDA measurements varied between 2 and 12 g/mn3. The
measured tar dewpoints are similar to the calculated values. In the temperature range of 25°C to
125°C the difference is at most 2°C. At a dewpoint of approximately 170°C the measured tar
dewpoint was 7°C to 14°C lower than the calculated tar dewpoint. A tar dewpoint of 219°C can
not be compared with the calculations, because the dewpoint is determined by tar compounds
that can not be measured with the SPA method.

Table 4.1 Comparison between the measured and calculated tar dewpoints.
 Label      Temperature   Calculated tar Measured Tar         Deviation                                     Total Tar
           tar condenser     dewpoint        dewpoint                                                     concentration
                [°C]            [°C]            [°C]             [°C]                                       [mg/mn3]
 10 Aug 1        60            56.2            57.1               0.9                                         3771
 10 Aug 2        60            55.6            56.5               0.9                                         3750
 11 Aug 1        80            68.8            69.6               0.8                                         5491
 11 Aug 2        80            72.9            72.2              -0.7                                         5185
 11 Aug 4       150            123.2           125.4              2.2                                         9936
 11 Aug 5       150            126.5           127.2              0.7                                         8996
 11 Aug 6       200            171.0           163.4             -7.6                                        12197
 11 Aug 7       200            179.4           165.9            -13.5                                        10649
 11 Aug 8       200            175.2           168.1             -7.1                                         9365
 12 Aug 1        20            27.4            26.2              -1.2                                         2019
 12 Aug 2        20            28.2            26.3              -2.2                                         2124
 12 Aug 3        20            27.3            26.3              -1.0                                         1833
 16 Aug 1       300                            219.0




24                                                                                                       ECN-C--05-026
The reliability of the tar dewpoint as measured at 219°C is questionable, as it became clear that
the fibre optics as delivered had a maximum operating temperature of 250°C. The high
temperature measurement started above a temperature of 250°C, and therefore the fibre optics
was probably thermally destroyed during the measurement, which leads to the conclusion that
for measurements above the 250°C the fibre optics should be replaced by other materials or
should be actively cooled.
                                 200
    Measured tar dewpoint [°C]




                                 150



                                 100



                                 50



                                   0
                                       0   50                 100                      150                     200
                                                 Calculated tar dewpoint [°C]

                                                Measured tar dewpoint         Y=X line
Figure 4.2 Comparison of the calculated tar dewpoint with the measured tar dewpoint. On the
           Y=X line the calculated tar dewpoints equal the measured tar dewpoints.

The shapes of the tar dewpoint curves are similar in the temperature range between 25°C and
170°C, as shown in Figure 4.3. The curves start (at high temperature) as almost horizontal
baselines and rise steeply near the dewpoints. The dewpoint curves for the high temperature
experiment has a different, less steep, slope. An explanation for the moderate slope can be the
low concentration of tar condensables in the product gas, or the damaging of the fibre optics
during the high temperature test of 300°C6.

The TDA can measure tar dewpoints at a low concentration of condensable tar. The amount of
tar condesables in product gas between 170°C to 180°C was estimated at 150 mg/mn3. Between
a temperature of 50°C to 60°C the amount of tar condensables is approximately 500 mg/mn3.
Despite such a large difference, the shape of the dewpoint curves is similar for the 170°C and
50°C measurements. Therefore, the TDA can measure below a concentration of condensable tar
of 150 mg/mn3. For dewpoint measurements in NG a lower limit of 10 ppmv (=100 mg/mn3 of
heavy tars (Pyrene-Coronene) is given. To improve the detection limit the TDA was operated
with a continuous gas flow.




6
    After the high temperature tests (last two tests) were finished, the signal lowered on recovery to 1000mV but could
    not be reduced any further. After cleaning the sensor cell with aceton the auto calibration resulted in an error
    message and caused the system to switch to the standby mode. Changing the fibre optics solved this problem and
    calibration could be performed.


ECN-C--05-026                                                                                                      25
                         2000

                         1800

                         1600

                         1400
    Signal Change [mV]




                         1200

                         1000

                         800

                         600

                         400

                         200

                            0
                                0               50              100                 150                  200          250              300
                                                                      Optical Surface Temperature [°C]


                                    11 Aug 1   11 Aug 2   11Aug 4       11 Aug 5      11 Aug 6      11 Aug 7   11 Aug 8     12 Aug 1
                                    12 Aug 2   12 Aug 3   10 Aug 1      10 Aug 2      16 Aug 1

Figure 4.3 Tar dewpoint curves.

As a conclusion, the proof of principle test was successful in the temperature window between
25°C and 170°C. The tar dewpoint sensor responds strongly to the tar condensables on the
optical surface. As soon as tar condenses on the optical surface the signal starts to increase, and
close to the tar dewpoint the signal increases exponentially. The shape of the dewpoint curve is
similar to the shape of HC dewpoint curves in natural gas. This means that the HC dewpoint
analyser is applicable for the detection of tar dewpoints in biomass product gases. The operating
temperature of the TDA is currently restricted by the temperature resistance of the fibre optics.


4.3                         Influence of water on measurement
The influence of water on the tar dewpoint measurements has been investigated with biomass
product gas from the WOB gasifier at ECN. Tar and dust free and water loaded product gas was
produced by drawing product gas through a filter for dust removal and an active carbon filter for
tar removal. The active carbon filter was operated above the water dewpoint. A normal
measurement cycle was performed in which the optical surface temperature was cooled down to
well below the water dewpoint. For comparison, the water dewpoint was measured at the inlet
of the sensor cell with an Optidew water dewpoint meter from Michell Instruments.

The water dewpoint measured with the TDA (64°C) was comparable to the water dewpoint
measured with the Optidew (63°C) analyser7. As soon as the optical surface temperature crossed
the water dewpoint temperature the signal increased exponentially, thus proving that the TDA
responds heavily on water in the product gas. Thus, for tar dewpoint measurements, water must
be removed from the product gas to such an extent that the water dewpoint is well below the tar
dewpoint.




7
    The water dewpoint was relatively high for product gases, because of the use of wet active carbon for tar removal.


26                                                                                                                        ECN-C--05-026
4.4    Gas conditioning section
The gas conditioning section performed well. Controlled and continuous tar, water and dust
removal was obtained. The aerosol filter downstream the tar condenser remained clean (white
colour), which indicates that no tar aerosols were formed in the tar condenser. The pressure drop
over the complete pre conditioning section was stable at 30 mbar, which indicates that no
significant fouling of pipelines and unit operations has occurred.


4.5    Pre design of a commercial tar dewpoint analyser
The TDA is ready for the next phase of development, the proof of concept, for measurements in
the temperature window between 25°C and 170°C. The proof of concept should be performed
with a pre design for a commercial TDA. For a commercial design the data recording and
measurement procedure of the modified HC dewpoint analyser should be automated. An
automated TDA will sample and measure continuously and will display the tar dewpoint
automatically after a measurement cycle is finished.

The present gas conditioning section is functioning well for tar dewpoint measurements between
25°C and 170°C. For continuous operation of the gas conditioning section the filter should be
replaced by a regenerative filter. The tar condenser should automatically remove the tar deposits
or water condensate. Finally the pump downstream the gas conditioning section should be
replaced with a pump that can be operated above the tar dewpoint, to preventfouling of the
pump with tar.

The layout for the pre design of a commercial TDA is given in Figure 4.4.

      dust removal            tar condenser           water removal           tar dewpoint
      (regenerative)                                  (optional)              analyser
                                                                              (optional)



                       water/tar condensate removal

Figure 4.4 Pre design of the gas conditioning section and TDA for tar dewpoint measurements
           between 25°C and 170°C.




ECN-C--05-026                                                                                 27
5.     MARKET AND COMPETING TECHNOLOGIES
An instrument that can measure the temperature of tar condensation does not exist yet. Several
tar measurement methods have been developed in the last 10 years. Most of them are based on
the sampling and off-line determination of the tar concentration and or composition in a
laboratory. Examples of these off-line techniques are the Solid Phase Adsorption method (SPA)
[8] and the tar protocol [9] developed by a European consortium with ECN as co-ordinator.

The competitor for the TDA is the FID tar-analyser [10]. The FID tar analyser is an online
method for the measurement of a total tar concentration. In the FID analyser de-dusted as well
as clean product gas are sent in sequence to a FID sensor. To obtain a clean gas, the product gas
passes a cold filter, on which tar condenses. The difference in signal between the raw and clean
gas from the FID represents the amount of tar that is removed by the filter. The advantage of the
FID analyser is the direct measurement of an amount of tar condensables, deposited on the
filter. The disadvantage is the detection limit of the analyser, which is strongly influenced
(lowered) by the presence of large concentrations of other hydrocarbons than tar such as
methane, ethane and ethylene in the product gas. Since most product gases contain a
considerable amount of light hydrocarbons, the FID instrument can mostly be applied in raw
product gas with high tar concentrations and is less applicable at low tar concentrations, in clean
product gas.

The market for the TDA in biomass gasification processes was identified and can be split in
three segments:
1. R&D groups
2. Stand–alone biomass gasification systems
3. Indirect co-combustion

R&D groups can use the TDA for the measurement of tar dewpoints at high and at low
temperature. Research activities in the field of biomass gasification concentrate on (1) the
optimisation of the gasifier and (2) the development of gas cleaning technologies for stand-
alone and co-combustion systems. Tar is an important issue for both activities. Reducing the tar
production inside the gasifier and the removal of tar in the gas cleaning, avoiding tar
condensation/fouling in the product gas cooler, piping, prime movers and other equipment, is a
key issue. For the prevention of tar related fouling in the cooler and hot gas filter, knowledge on
tar condensation at a temperature between 150°C and 350°C is essential. For the protection of
the prime mover (e.g. gas engine), using tar removal technologies, knowledge of tar
condensation at low temperature is needed (typically below 25°C). Therefore R&D groups can
benefit from the tar dewpoint sensor in the full temperature window between 0 and 350°C in the
development of the product gas cooler, hot gas filter and NH3 and tar removal technologies.

Stand-alone biomass gasification systems can use the TDA in the prevention of tar-related
fouling. Stand-stills due to fouling have a direct impact on the pay back time of the installation,
because of the obvious loss of income and extra costs to be made for the replacement or
cleaning of equipment. The total costs for tar-related fouling have been quantified by
Van der Drift [11] for a 12 MWth biomass gasification system at 1.2-2.4 €ct/kWhe. The TDA
can be used for protection of the product gas cooler, hot gas filter, NH3 scrubber and prime
mover against tar related fouling. The temperature window for the application of the TDA is
therefore identical to the R&D segment. Due to the high investment costs for a gas engine or a
gas turbine, the protection of these prime movers against tar related fouling is a priority.
Therefore, the application of the TDA at low temperature between 0°C-100°C is most
important.

Co-combustion systems using a biomass gasifier installed upstream a coal boiler can also
benefit from the TDA in the prevention of tar related fouling in the cooler. Indirect co-


28                                                                                ECN-C--05-026
combustion systems, like the power plant AMER in Geertruidenberg, use a gasifier for the
conversion of biomass. Downstream the gasifier the gas is cooled down to intermediate
temperatures of approximately 400°C, de-dusted with a cyclone, and than fired on the coal
boiler to replace a part of the coal. Fouling of the cooler is an important risk in which tar
condensation plays an important role, as was concluded in [2]. In the prevention of cooler
fouling, the TDA should be able to measure tar condensation at high temperature between
180°C and 350°C.

The markets as defined above can be separated in a short term and long term market. R&D
groups and co-combustion plants are potential end-users of the TDA for the short-term market.
Several R&D institutes and universities in the world can benefit from a TDA, because it will
give a researcher and an engineer more insight in the tar condensation properties. For co-
combustion, a few commercial installations like Lahti, Amer and Ruien are already running.
This market is still increasing and is therefore a short-term market. The market for the
application in stand-alone gasification systems is expected to increase in time and is therefore
identified as a long-term market. An example of a stand-alone gasification demo plant is the
Guessing installation in Austria.

In terms of market perspectives the proof of principle was successful for the application in R&D
groups and stand-alone biomass gasification systems, so for measurements at low temperature.
The temperature window is not high enough for the application in indirect co-combustion
systems but can potentially be increased with a few modifications to the TDA.




ECN-C--05-026                                                                                29
6.     CONCLUSION
In this project a tar dewpoint analyser is developed for the on-line measurement of tar
dewpoints in biomass product gases. The analyser consists of a tar dewpoint sensor and a gas
conditioning section. The sensor is the key element of the instrument as it physically measures
the temperature of tar condensation, the tar dewpoint. The sensor is a modified version of the
existing hydrocarbon dewpoint sensor that is normally applied for natural gas characterisation.
The gas conditioning section has been designed and built to protect the sensor against fouling
and corrosion.

For the application of the sensor three commercial relevant operating conditions were specified
for the tar dewpoint sensor and the gas conditioning section. For the application in raw product
gas, case 1 in the objective, the sensor must be modified for operating at a high temperature
(60°C-350°C). Downstream a water-based tar removal system, case 2 in the objective, the tar
dewpoint is comparable to the water dewpoint (20°C-60°C). Therefore water must be removed
in the gas conditioning section to prevent interference of the measurement by water.
Downstream advanced tar removal methods, case 3 in the objective, the tar dewpoint is far
below the water dewpoint (<60°C) and water should be extensively removed in the gas
conditioning section.

The original fibre-optics were replaced by a high-temperature resistant (250°C) system. The
synthetic material in the original sensor was replaced by ceramics. The internal diameter of
piping was increased to prevent blockage by tar and the sensor cell was placed in an oven to
avoid cold spots in the sensor. The gas conditioning consists of a filter, tar condenser and water
removal section. The filter removes dust and alkali metals in the system. The tar condenser is at
normal operation not active but protects the water removal section or the dewpoint sensor
against heavy tar deposits. The water removal section dries the product gas above the tar
dewpoint, to prevent disturbances of the measurements by the presence of water (cases 2 and 3).
A maintenance period of 3 months is defined as a minimum period for the gas conditioning
section.

In a preliminary test the performance of the water removal section, the tar condenser and sensor
recovery were investigated. The water removal section could reduce the water dewpoint with
40°C. The water removal section must be operated above the tar dewpoint to prevent tar
condensation. The maximum operating temperature is determined by the thermal resistance of
the materials, which is 80°C. The tar condenser could be operated stably and removes tar
without the formation of tar aerosols. The temperature of sensor recovery should be 30°C to
50°C above the tar dewpoint. Irreversible fouling of the sensor with polymerised tar was
identified as a potential risk for tar dewpoint measurements at high temperature (>280°C).

The operational proof of principle test with the dewpoint analyser (sensor and gas conditioning
section) was successful. The tar dewpoint sensor responds strongly to tar condensables and the
gas conditioning section protects successfully the dewpoint sensor against fouling and
corrosion. The shape of the dewpoint curve is similar to the shape of hydrocarbon (HC)
dewpoint curves in natural gas. This means that the HC dewpoint analyser is applicable for the
detection of tar dewpoints in biomass product gases from a fluidised bed gasifier. The TDA has
measured tar dewpoints at a concentration of tar condensables between 150 and 500 mg/mn3.

The TDA accurately measured tar dewpoints in a temperature window of 25°C to 170°C. The
operating temperature of the TDA is limited by the thermal resistance of the fibre optics.
Therefore, the maximum tar dewpoint is limited to 200°C. This means that the current TDA was
successfully tested for the application downstream tar removal technologies, cases 2 and 3. For



30                                                                               ECN-C--05-026
this temperature window the instrument is ready for the next phase of development - the proof
of concept.

The proof of concept tests should be performed with a pre commercial design of the dewpoint
analyser, measuring in clean product gas downstream a tar removal unit. The measurement
procedure and gas conditioning section should be automated. An automated TDA will sample
and measure continuously and will display the tar dewpoint automatically after a measurement
cycle is finished. For a continuous operation of the gas conditioning section, the static filter
should be replaced by a regenerative filter. The tar condenser should automatically remove the
tar deposits and/or water condensate. Finally the pump downstream the gas conditioning section
should be replaced with a pump that can be operated above the tar dewpoint. For high
temperature measurements the fibre optics should be cooled or replaced by high temperature
resistant fibre optics.

The market for the TDA in biomass gasification processes was identified and can be split in
three segments:
1. R&D groups
2. Stand–alone biomass gasification systems
3. Indirect co-combustion
The markets can be divided in a short term and long term market. R&D groups and co-
combustion installations are potential end-users of the TDA for the short-term market. The
market for the application in stand-alone gasification systems is expected to expand in time and
is therefore identified as a long-term market. The current tar dewpoint instrument, with a
temperature window between 25°C and 170°C, is applicable in the R&D and stand-alone
markets. For the co-combustion market the sensor must be modified for the operation at high
temperature.




ECN-C--05-026                                                                                31
REFERENCES
 1. J. P. A. Neeft: Teren uit thermische conversie van biomassa en reststromen. Definities,
    vorming, eigenschappen, en bemonstering en analyse, Petten, ECN, ECN-C--99-102, 58
    p. (2000).
 2. A. van der Drift and J. R. Pels: Product gas cooling and ash removal in biomass
    gasification, ECN, Petten, The Netherlands, ECN-report: ECN-C-04-077, 42 p. (2004).
 3. D. Dayton: A Review of the Literature on Catalytic Biomass Tar Destruction, NREL/TP-
    510-32815 (2002).
 4. H. Boerrigter, S. V. B. van Paasen, P. C. A. Bergman, J. W. Könemann, R. Emmen,
    A.Wijnands: OLGA tar removal technology, ECN, Petten, the Netherlands, ECN-C-05-
    009 (2005).
 5. L. P. L. M. Rabou, S. V. B. van Paasen, S. M. Willemsen, E. de Kant and T. Smit:
    Onderzoek naar het betrouwbaar maken van teer en ammoniakverwijdering uit
    biomassastookgas, Petten, ECN, ECN-CX--04-048, 50 p. (2004).
 6. L. P. L. M. Rabou and A. van der Drift: Optimalisatie gasreiniging (confidential), ECN,
    Petten, ECN-CX-02-113, 52 p. (2002).
 7. T. A. Milne, N. Abatzoglou and R. J. Evans: Biomass gasifier tars, Their nature,
    formation and conversion . 1-68 (1998).
 8. C. Brage, Qizhuang Yu, Guanxing Chen, C. Rosén, T. Liliedahl and K. Sjöström:
    Application of solid-phase adsorption (SPA) to monitoring evolution of biomass tar from
    different types of gasifiers. In: Biomass Gasification & Pyrolysis. State of the art and
    future prospects (Eds. M. Kaltschmitt and A. V. Bridgwater), pp. 218-227, CPL Press,
    Newbury (UK) (1997).
 9. S. V. B. van Paasen, J. H. A. Kiel, J. P. A. Neeft, H. A. M. Knoef, G. J. Buffinga, U.
    Zielke, K. Sjostrom, C. Brage, P. Hasler, P. A. Simell, M. Suomalainen, M. A. Dorrington
    and L. Thomas: Guideline for sampling and analysis of tar and particles in biomass
    producer gases, Petten, ECN, ECN-C--02-090, 94 p. (2002).
10. B. Staiger, L. Wiese, R. Berger and K. R. G. Hein: Investigation of existing gasifier and
    gas cleaning technologies with an online-tar measuring system, pp. 789-792, Rome, Italy
    (2004).
11. A. van der Drift, H. F. de Kant and J. B. Rajani: Commercialisation BIVKIN-based
    gasification technology, non-confidential version, Petten, ECN, ECN-C-00-080, 56 p.
    (2000).




32                                                                           ECN-C--05-026

				
DOCUMENT INFO