![[BioMatNet Database - FAIR Program]](../images/Fair.gif) |
FAIR-CT97-3571
BIOMIS Reduction of fouling, slagging and corrosion characteristics of Miscanthus for power and heat generation using biotechnology |
 |
Contract No: | FAIR-CT97-3571 |
| Date Prepared: | April 2001, October 2003 | |
| Source: | Progress Report | |
| Final Report - Abstract | ||
| Final Synthesis Report, Executive Summary |
Biomass from the bioenergy crop Miscanthus sinensis (2n=38) has in comparison to wood a relatively poor quality in relation to power and heat generation due to relatively high concentrations of minerals. A major aim of the current project was to improve the combustion quality of this crop by means of modern (bio)technological techniques. To this end adequate technological tools have been developed to study the combustion quality of biomass per se and biotechnological tools have been developed to analyse the genetic variation for combustion-related traits. The respective tools are a laboratory-scale test rig, and an integrated molecular map with two types of DNA markers, being RAPDs and AFLPs.
The test rig allows pilot studies of combustion quality using small amounts of biomass (10kg). In the test rig the combustion gases are led over a water-cooled metal sample, to simulate a heat exchanger in a biomass boiler. The duration of a test is two weeks. A total of seven tests were conducted with various biomass samples, including miscanthus from Denmark and Spain, Austrian M. x giganteus and beechwood for reference. Main results were that miscanthus with a higher Si/K and Ca/K ratio showed almost no tendency towards slagging, and that miscanthus with higher amounts of chlorine led to low-temperature pit corrosion. No relation between sample and deposit formation could be found, and also no relation between sample and other corrosion mechanisms could be established. Miscanthus feedstocks did cause significantly more deposits than beechwood. These deposits however showed no dependency on temperature.
Chemical equilibrium models were used to validate the test results of biomass samples and to get a better understanding of the impact of mineral composition on combustion. The models proved to be an effective tool for evaluating the influence of the chemical composition of the fuels on the formation and chemistry of ashes formed upon combustion of different fuel samples in the test rig. The predicted release of K from the fuel bed during combustion as well as the predicted slagging characteristics of the residual ash were in agreement with the observations. Lowering the K and Cl contents of biomass is very relevant. The fouling of ash on heat recovery surfaces, on the other hand, was not predictable by chemistry alone. However, the relatively high miscanthus ash deposits turned out to be porous and easily removable.
The integrated map, the second tool, consisted of 19 linkage groups presumably representing all or nearly all chromosomes. The linkage groups ranged in size from 26cM to 134 cM, giving together a map size of 1259 cM. The map was used to study the genetic variation for combustion-related traits, being the contents of K, Cl, Ca, S, N, and P. The genotypic data were obtained from a 2-year evaluation of the mapping population in Spain and Denmark. In addition several yield and various shoot fractions were evaluated. The trials had a randomised block design with two replicates. The experimental units were small plots of three plants. The overall statistical analyses showed a considerable degree of heritable variation for the contents of K, Cl, N, and P as well as for yield and dry matter of the stem fractions studied.
The marker analyses of genotypic variation revealed in total twelve so-called quantitative trait loci (QTLs) with a LOD-value above 2.5 for the chemical constituents of stems. The number of QTLs detected per trait differed from three for K, Ca, and Cl to nil for S, whereas two were found for P and one for N. QTLs were found on nine different linkage groups. The strongest QTL was found on linkage group I. being a putative gene for potassium, which co-segregated with a QTL for chlorine content. It had an LOD-value of 6.6 and accounted for 37% of the variation among genotypes for stem potassium content. Yield and shoot dry matter components showed in general less striking QTLs than most chemical constituents of stems, despite higher heritabilities. In total eight QTLs were found distributed over six linkage groups: only two yield-related QTL.s had a LOD-value above 3. On the other hand three strong QTLs with a LOD-value above 3 were found for flowering; one inkage group had a value of 7.6, accounting for 24 % the variation among genotypes. The results showed that there is ample scope for using molecular markers in breeding, for low K and Cl contents in particular.
The project further opened the door for a second way for genetic improvement of the combustion quality of miscanthus, being genetic transformation. The primary aim was to develop a transformation protocol using two different approaches. A successful protocol was finally obtained for the panicle gun approach.
The current project has shown that improvement of the fuel quality in M. sinensis is possible by conventional breeding with or without marker selection and/or by genetic transformation. The economic analyses showed that there is a very sound economic basis for breeding for fuel quality. The development of high quality cultivars is only expected if energy companies pay more for high-quality biomass.
Objectives
The main aims of this shared cost project were:
Molecular Map of Miscanthus sinensis
A molecular map of M. sinelisis (2n=38) was generated in order to be able to analyse the genetic variation for combustion-related traits. The population used was an offspring of a cross between two full sibs, being P1.1 and P1.7. derived from a cross between two heterozygous plants. i.e. MS-90-2 and MS-88-11O. In total 505 polymorphic markers RAPD and AFLP markers were used for mapping. In part the linkage phase of markers were established from the genotyping data of parents and grandparents of the mapping population. Most dominant uni-parental markers from the male and female parent, together 271 markers, could be mapped easily. The integration of the female and male maps, however, was more cumbersome because the dominant bi-parental markers are not very well suited as anchor points for map integration. The high number of linkage groups made this task even more difficult, Nevertheless the map construction resulted in a map with 19 linkage groups presumably representing all or nearly all chromosomes. The linkage groups ranged in size from 26cM to 134 cM, giving together a map of 1259 cM. A rough estimate of the genome coverage with molecular markers is 66%.
Variation for combustion-related traits
Improvement of the combustion quality of miscanthus biomass was a major poject aim. Combustion quality comprises all characteristics of biomass having an effect on its suitability for power and heat generation, in particular their effect on the formation ashes, slags and corrosion during the combustion process. To enable genetic improvement the magnitude of the available genetic variation for the relevant traits, being the contents of K, Cl, N, F, Ca, and S in biomass, was studied in field trials during two growing seasons in Spain and Denmark. In addition yield, dry-matter distribution over shoot fractions and the onset of flowering were determined. The environment (= combination year and location) had a considerable influence on the performance of population A. In Spain the chlorine contents of stems were much higher than in Denmark. The same held true for the contents of potassium and phosphorus. The opposite held true for nitrogen content. Genotype accounted for a moderately high proportion of variation for the contents of chlorine, nitrogen, calcium and potassium, whereas the factor genotype had almost no influence on the variation for sulphur and phosphorous. The mean dry-matter production of genotypes was on average 925 kg/m2 and was considerably higher in the second production year. The mean production was fairly low but this was probably due to starting problems and incomplete soil coverage. The stem fraction in the whole crop covered about 87 % of the crop yield with a range of 8.5%-units. The other fraction, i.e. leaves and panicles were fairly insignificant. There was further a very significant variation for date of flowering, on average more than 40 days. Genotypes differed highly in number of shoots per plant. All yield-related traits, i.e. total shoot drymatter weight, and its major components and flowering showed fairly high heritabilities.
Genetic Analyses of Variation
The analyses of genotypic variation revealed in total twelve so-called quantitative trait loci (QTLs) with a LOD-value above 2.5 for stem chemical constituents. The QTLs detected, however, differed strongly in likelihood. The number of QTLs detected per trait differed from three for K, Ca, and Cl to nil for 5, whereas two were found for P and one for '~. QTLs were found on nine different linkage groups. The strongest QTL, was found on linkage group 1, being a putative gene for potassium which co-segregated with a QTL for chlorine content. It had a LOD-value of 6.6 and accounted for 37 % of the variation among genotypes for stem potassium content. Yield and shoot dry-matter components showed in general less striking QTLs than most chemical constituents of stems, despite higher heritabilities. In total eight QTLs were found distributed over six linkage groups; only two yield-related QTLs had a LOD-value above 3. On the other hand three strong QTLs with a LOD-value above 3 were found for flowering; one on linkage group had a value of 7.6, accounting for 24 % the variation among genotypes.
Transformation
One of the project goals was to develop a transformation protocol for miscanthus. Therefore transformation experiments were initiated using the Miscanthus sinensis clones: MS88-110, MSS8-111 and MS9O-2. These clones showed good regeneration capacity in tissue culture. Plasmids used for transformation contain the Bialaphos resistance gene (BAR) encoding phosphinothricin acetyltransferase and the beta-glucoronidase gene (GUS) to select for successful transformation events. In the set of plasmids used the genes were driven by various promoters. Two different routes were followed to develop a transformation protocol. The first one, being Agrobacterium transformation was carried out using different Agrobacterium strains, explant types. co-cultivation periods and pre-treatment of the callus. A few GUS-positive calli were obtained but none of the calli could be regenerated into transgenic shoots. The second route, transformation by particle bombardment was more successful. This technique was mainly carried out on embryogenic callus. The pretreatment of the callus and the bombardment conditions were varied to obtain an optimal protocol. Polymerase chain reaction (PCR) analysis showed that callus had obtained the foreign DNA. The calli, however, showed no GUS expression. Nine independent calli gave shoot regeneration. The shoots were cloned and induced to root. The transgenic nature of the individual plants was confirmed by PCR and Southern analysis. So,it was concluded that transformation of M. sinensis by particle bombardment is possible.
Test rig and Combustion tests
An existing test rig was modified to allow for two-week tests with only 10 kg of material. For this a new feeding device was developed, in which biomass was ground to millimeter sized particles, and subsequently compressed to soft pellets. These were hydraulically fed to a rotator, which scraped off the pellets, after which these would fall into the externally heated reactor, and be rapidly combusted. The combustion gases were led over a water-cooled metal sample to simulate a heat exchanger in a biomass boiler. A total of seven tests were conducted with various biomass samples, including miscanthus from Denmark and Spain. Austrian M. xgiganteus, and beechwood for reference. Main results were that miscanthus with a higher Si/K and Ca/K ratio showed almost no tendency towards slagging and that miscanthus with higher amounts of chlorine led to low-temperature pit corrosion. No relation between genotype and deposit formation could be found, and also no relation between genotype and other corrosion mechanisms could be established. Miscanthus feedstocks did cause significantly more deposits than beechwood. These deposits, however showed no dependency on temperature.
Use of chemical models for combustion quality
Models were used to validate the test results of biomass samples. The evaluations based on these models, so-called chemical equilibrium equations, showed that the models are an effective tool for evaluating the influence of the chemical composition of the fuels on the formation and chemistry of ash chemistry formed upon combustion of different fuel samples in the test rig. The calculations showed the relevance of lowering the K and Cl contents of biomass. The predicted release of K from the fuel bed during combustion as well as the predicted slagging characteristics of the residual ash were in agreement with the observations. The fouling of ash on heat recovery surfaces, on the other hand, was not predictable by chemistry alone. However, the relatively high miscanthus ash deposits turned out to be porous and easily removable; therefore having only a minor effect on the combustion quality.
Economic analysis
An evaluation of the direct benefits of R&D into genetic improvements of the miscanthus fuel quality, being reduced investment and avoided maintenance costs, has been made. In addition R&D into fuel quality could have indirect socio-economic advantages, such as reduction of emission of green house gasses, reduced depletion of fossil fuels, fuel diversification, alternatives for the agricultural sector, increased employment, and enhanced options for nature and recreation. Most social objectives, however, are achievable through alternative renewable energy sources or import of bio-fuels, except for the perspective offered to the agricultural sector. The total incremental use of biomass fuels in the EUI5 in the period from 1995 to 2010 is estimated to be 90 to 98) Mtoe/year. The demand increase equals to a growth by a factor of three over a period 15 years. Energy crops are expected to supply half of the demand, being 45 Mtoe/year. Energy crops are supposed to deliver yearly 27 Mtoe of solid fuel and 18 Mtoe liquid fuel. The respective areas of land required to meet this demand are 6.3 Mha and 5.6 Mha. This is more than the available area of land for energy crops in EUI5 (7.6Mha; 10% of available arable land). For the economic analysis of R&D into miscanthus breeding for improved fuel quality (1 Meuro/year for a period of ten years) it was assumed that the potential area for miscanthus is 6.3 Mha (27 Mtoelyear). Other assumptions were that about half of the crops will be utilised with technologies for combined heat and power generation (CHP). Non-CHP technologies will be used for valorisation of the other half of the crops. Quantification data on the cost reductions possible by growing improved miscanthus was obtained from a comparative study of the utilisation of wood and straw in a Non-CHP heat plant with 2MW boiler capacity per annum and 100% conversion efficiency. The further assumptions were that un-improved miscanthus biomass resembles straw, and improved biomass wood. The potential cost savings due to crop improvement were estimated 3 euro per MWh for (annualised initial) capital costs and one euro per MWh for the operational costs. In case of a successful project, the cost savings for energy generation are 44 Meuro/Mtoe. For the anticipated situation for EU 15 with an annual production 27 Mtoe of solid energy crops in the form of Miscanthus, genetic improvement of fuel quality would imply an annual cost reduction of over 1 billion euro. The cost reductions achievable easily justify a 10-year breeding program of on million euro per year into improvement of the combustion quality of miscanthus biomass. This already holds true for a small share of miscanthus in the EU15 market of energy crops for solid fuels (<7000 ha with yield of 15 t/ha/year).
Conclusions
The subject of study in the current project was M. sinensis and not M. xgiganteus, the most widely used miscanthus form. Nevertheless it is thought that the project findings can be best implemented through improving M. sinensis as such. Although the field trials were not optimal for yield evaluations, they clearly showed that with an adapted crop management the species has a high potential as a productive producer of a high quality biomass, having many advantages in comparison to M. xgiganteus from a breeding point of view. The advantages are seed fertility, self-incompatibility, and a high multiplication rate; characteristics allowing, in principle, the development and commercialisation of seed-propagated varieties. As a result of the current project improvement of this crop is now possible by conventional breeding with or without marker selection and/or by genetic transformation. The economic analyses showed that there is a very sound economic basis for a breeding programme aiming at improvement of the combustion quality of miscanthus per se. However, development of 'high quality' cultivars will only happen if energy companies are prepared to pay a higher price for high-quality crops. As long as there is no certainty with respect to this matter it is unlikely that the long-term investments and efforts needed for miscanthus breeding will happen. It is up to the EU, national governments and/or the energy companies to find ways to make breeding of productive and high-quality energy crops attractive.
Objectives
The aim of this project is to reduce significantly cleaning, maintenance and replacement costs for expensive heat exchangers (piping) in the thermal conversion processes (combustion, gasification and pyrolysis) of Miscanthus biomass by genetic improvement of its fouling, slagging and corrosion characteristics. Molecular breeding tools will be developed in order to do this. The objective is to minimise concentrations of elements having an adverse effect on corrosion (Cl, K, and S) and on slagging and fouling (K, Ca, and P). The project also aims to further integrate the whole bioenergy chain, i.e. the agricultural and energy sector, and to demonstrate the usefulness and perspective of biotechnology for structural and cost- effective improvement of biomass fuel characteristics.
Activities
The work consists of four research tasks.
The variation for each trait will be dissected using the molecular map of Miscanthus and the DNA marker profile of the individual Genotypes. The final outcome will be the mapping of loci for quality characteristics and some biomass production characteristics. The mapping will enable indirect selection for these traits using a molecular markers flanking such, so-called, quantitative trait loci (QTLs). The second biotechnological approach for improvement of Miscanthus envisaged during this project is the development of a protocol for genetic transformation of Miscanthus.
The research on transformation is focussed on two different methods, being particle bombardment or Agrobacterium-mediated transformation. The starting materials used for transformation are either leaf explants or shoot apices of genotypes or embryogenic callus. The final aim is to transform Miscanthus with gene constructs to improve cold tolerance or quality. In addition to the elemental analyses, tests on fouling, slagging and corrosion characteristics will be carried out using special test rigs, consisting of a small combustion chamber and controlled cooled surfaces on which fouling components can deposit and corrosion may take place. Slagging characteristics will be determined by ash melting analyses.
The data from these activities will be integrated into an overview covering the whole bioenergy chain of Miscanthus, matching the results from the agricultural partners with those of the energy partners. The evaluation includes a financial/economic assessment in order to evaluate the costs needed for improvement of Miscanthus using biotechnology and the resulting savings in the energy sector. The costs of the overall chain will also be estimated.
Progress
Genotyping of population A and its parents and grandparents included AFLP, RFLP, SSR and RAPD analyses and resulted in a large amount of data which will form the basis for the planned inheritance studies. However, the mapping data of the partners has yet to be combined to get one final map. A reason for the delay in map construction is that genotyping continues. The RFLP and SSR analyses, that are of particular importance in establishing synteny-relationships between Miscanthus and other gramineous species, have been started. In the mean tie the field evaluation of population A as well as the corresponding chemical analyses is in progress. The results of first- harvest from Spain are available for genetic analysis.
The investigations to develop a protocol for transformation of Miscanthus were continued and were successful. The biolistic approach resulted in transgenic plants resistant to bialophos. The test rig, for laboratory-scale tests of the combustion quality of Miscanthus biomass were further improved and its use for the duration tests is expected. Samples from the Danish trial will be made available soon. In parallel extensive modelling studies were done to study the impact of changes in the chemical composition of biomass.
Achievements
The current project is quite ambitious. Nevertheless a steady progress has been achieved in all major research tasks of the project, although in some probably somewhat slower than envisaged. Highlights of the research are an effective transformation protocol using a particle gun, a test rig for laboratory-scale duration tests for biomass combustion and the insights in the implications of changes in the chemical composition on the melting behaviour of ashes from Miscanthus biomass.
Chemical equilibrium calculations clearly revealed that the most relevant elements for slagging and fouling on the grate and in the furnace are primarily K and secondly Ca. A high concentration of K and low concentration of Ca increases slagging and fouling while a low concentration K and high concentration of Ca decreases this tendency. Consequently, in order to improve the slagging and fouling tendencies of Miscanthus the main breeding objective has to be to decrease the amount of K in the plants. However, also a higher Ca concentration could have a positive effect on the slagging and fouling properties of the fuel.
The data collection for the inheritance study is on track and resulted in a nearly complete data set summarising the performance of population A and its parents and grandparents during the first-evaluation year in Spain showing considerable genotype- to-genotype variation for most traits. In parallel the material also has been genotyped with various types of markers. Genetic mapping, still needs some extra genotyping efforts before integration of the genotyping data of both partners. All ingredients are than available for analysis of the variation for combustion-related traits.
Future activities
The research activities for the final year will be as planned. The main actions are to:
At the end of the project a workshop on improvement of the combustion-quality of Miscanthus biomass will be held where all partners will present their results.
© Copyright 2006 Policy Statements
Updated
by CPL Press:
03/07/2007
- biomatnet@biomatnet.org
 |
 |
 |
News |
 |
Events |
|
|