The review of the proposed biomass conversion of the Jasper Power Plant by Dr. Shaddix

Review of Twisted Oak Proposal and Associated Literature

Christopher R. Shaddix, PhD

Consultant to Jasper, IN Utilities Board

July 5, 2011

Summary of proposal:

Twisted Oak proposes to repower the existing 14.5 MW stoker grate boiler, previously operated on a local coal source, as a 15 MW biomass boiler utilizing locally grown Miscanthus.1 In addition, a 65 MW natural gas combined cycle power plant will be installed with shared use of a new 32 MW steam generator. The project will involve installation of a new biomass receiving and processing facility, a new gas turbine generator, a new steam condenser, upgrades to existing boiler emission control equipment, upgrades to the cooling towers, a new diesel generator set, a new generating plant substation, and upgrades to the 69 kV electrical interconnection to the wholesale electric grid. Furthermore, life extension improvements will be performed on the existing boiler.

Choice of Miscanthus as boiler fuel

Of all of the energy crops possible in Indiana, Miscanthus appears to be the best option for use as a boiler fuel: it has good mass and energy yields, requires little fertilizer, is simple to harvest, has low moisture when harvested in early spring, and has the lowest ash, nitrogen, and chlorine content of any herbaceous biomass crop.2?7 Extensive cultivation tests and boiler tests have been conducted with Miscanthus in Europe. In contrast, U.S. experience is much weaker, despite favorable reports of mass yield in several U.S. locations.2 Miscanthus originated from warm weather locations in Asia and Africa, but has shown good overwintering properties in many European locations. However, first winter mortality can be high, and the establishment costs of Miscanthus, through rhizome planting, are also high.8 Research is being conducted on establishing Miscanthus through stem planting.8 Competition from weeds can be a major problem during the 1? to 3?year establishment period of the Miscanthus, but recent success with the well?known and low?cost herbicide atrazine is promising.9

Pilot?scale combustion tests and small boiler tests with Miscanthus show much lower NOx, CO, and dust emissions than when using other herbaceous fuels.3?5,10 In general, combustor performance and emissions with Miscanthus are similar to performance with clean wood.3?5 Details regarding combustor performance and emissions performance will be further discussed below.

Project economics

It is very difficult to generate power for profit in the wholesale power market when using a cultivated biomass fuel source, because such fuel is likely to cost on the order of $60?$80/ton – especially after accounting for storage losses that occur during storage. Oak Ridge National Laboratory has reported 30?50% mass losses upon overwintering the Miscanthus, which is advisable to allow for harvesting of dry

Miscanthus in the early spring.2 Because of the challenging economics of dedicated energy crop power generation, no one in the U.S. (to my knowledge) has initiated operation of a similar biopower project using a dedicated energy crop. Twisted Oak has suggested that cost projections (with an unspecified Miscanthus feedstock cost) show the biopower cost to be below local solar photovoltaic power costs and similar to local windpower costs. This is believable, in part because Indiana does not have a favorable climate for either solar power or windpower. The implication given by Twisted Oak is that the biopower will be purchased at whatever cost, because of the current voluntary requirement for renewable energy purchases in Indiana and/or by marketing the power to neighboring states that have Renewable Portfolio Standards (RPS’s) in place.11 However, with the hybrid steam turbine design, utilizing steam generated by both the biomass boiler and the natural gas?fueled turbine exhaust, it is unclear whether there are regulatory or legal issues associated with defining a given percentage of the steam turbine output as renewable biopower.

Boiler fuel feeding

A significant area of concern is with the manner in which Miscanthus will be introduced and burned within the boiler. Because of its low density (both in terms of mass density and energy density), the volumetric feed rate of biomass needs to be much greater than that of coal, the fuel for which the boiler was originally designed. Furthermore, non?pelletized herbaceous biomass, such as Miscanthus straw, has significantly lower density than wood, and boiler derates have been reported to be necessary when retrofitting a stoker grate boiler designed for wood firing for herbaceous biomass.12 Retrofitting a coal?fired grate to firing on herbaceous biomass is an even more daunting challenge. Twisted Oak proposes making either briquettes or pellets out of 20?30% of the Miscanthus for feeding onto the grate and then feeding the remainder of the Miscanthus as milled straw over the grate. This feed rate of Miscanthus pellets onto the grate will probably mimic the volumetric coverage of the grate with coal feeding.

The planned operation of the boiler with a high feed rate of low?density air?blown feed above the grate raises questions with respect to air flow distribution, heat release profile, and unburnt fuel carryover (leading to problems of CO, tar, and dioxin/furan emissions, in addition to efficiency loss). Twisted Oak indicates that they expect to develop an understanding of the allowable fuel feed rates through boiler modeling and analysis with a company with expertise in stoker fuel feeding, such as Detroit Stoker.11 Further, Twisted Oak suggests that they actually expect the boiler to be derated when operating on Miscanthus. They plan on meeting the 600 psi steam pressure setpoint for the steam generator, but expect that the level of superheat will be reduced when operating on Miscanthus. For this reason, the heat?recovery steam generator (HRSG) is being designed with a separate natural gas feed to allow for additional heat input to keep the steam turbine operating at capacity.11 Twisted Oak has not suggested that it has spoken with boiler operators in Europe concerning their experience with feeding Miscanthus and with managing Miscanthus fuel distribution in grate?fired boilers.

Boiler slagging and fouling

Biomass fuels are notorious for their tendency to produce extensive slagging and fouling in combustion systems. 13,14 The alkali content, chlorine content, and sulfur content all play roles in slagging and fouling

behavior, in complicated ways. The low overall ash content of Miscanthus is promising in this regard, as it means the amount of inorganic material that the boiler needs to deal with is lower than for other herbaceous fuels, such as switchgrass. Brunner et al.12 have found particle deposition rates to be several times greater when burning Miscanthus than when burning clean wood. The Jasper boiler has a history of excessive slag formation, which can only be expected to worsen when firing biomass, particularly herbaceous biomass such as Miscanthus. In response to this concern, Twisted Oak is indicating that they intend to install additional soot blowers, sonic horns, and screens within the steam tube passages and are anticipating including ~ 36 hrs per month of boiler down time to perform maintenance and cleaning of boiler surfaces.11 In Europe, biomass boilers typically operate with flue gas recirculation (FGR) both below and above the grate in order to lower the peak temperature in the boiler and to minimize slag formation.12,15 Twisted Oak is considering the use of FGR, but is awaiting the results of boiler analysis and modeling before deciding how or if to implement this.11

Boiler emissions and emission control

A large part of the concern with the proposed Twisted Oak repowering project is the emissions that can be expected or which may be possible when operating the stoker grate boiler on Miscanthus or other biomass sources. In comparison to coal, biomass has a lower sulfur content and therefore generally produces lower SO2 emissions. However, other emissions, such as NOx, may be comparable to those from coal, and potential difficulties in fuel feeding or maintaining a consistent moisture content in the biomass feed lead to concern over emissions of fugitive hydrocarbons, CO, and dioxins and furans. In addition, the alkali content of biomass leads to concerns over emissions of ultrafine particles that can travel deeply into lung tissue and potentially have greater health impact than the larger particles that are typically emitted from coal boilers.

Particulates

As mentioned above, an emission concern from the combustion of biomass is particulates, particularly ultrafine particles that are generated when alkali metals (especially potassium) are volatilized and then condense as the combustion product gases cool towards the back end of the boiler. Herbaceous biomass sources generally produce higher concentrations of ultrafine particles than clean wood because they have higher concentrations of potassium and sodium. These particles typically undergo aggregation within the boiler backpass and can deposit onto boiler surfaces or onto larger particles. The formation and emission rates of these particles depend on many factors, including the fuel composition, boiler operating temperature, and the presence or absence of larger particles to help ‘scrub’ the small particles out of the system. Miscanthus, with its very low ash levels (typ. 2?3%) and low chlorine levels (typ. 0.1%), has a much lower tendency to form particles in the boiler flue gas than other herbaceous biomass sources, but it also produces lower concentrations of larger particles that help scrub out the ultrafine particles. Boiler and pilot?scale furnace tests suggest that Miscanthus has a similar, but slightly higher, tendency of forming ultrafine particles as clean wood.3

The most significant aspect influencing the emission rate of particles from the boiler is the particulate control device(s) that are being employed. Previously, when operated on coal, the Jasper boiler relied on

a multiclone coarse particle collector followed by an electrostatic precipitator (ESP). ESPs can be quite effective at capturing particles, when properly tuned and when the particle population that is entering the ESP shows consistent properties (electrical resistivity, size distribution, etc.). ESPs are not well suited to operations in which the fuel feeding or other aspects of the boiler operation are inconsistent, leading to variable particle properties. For this reason, Twisted Oak is proposing to replace the existing ESP with baghouses. Baghouses are insensitive to variations in particle loading and particle properties. Baghouses are also generally superior to ESPs in capturing ultrafine particles, though there are many variables in both baghouse design and operation that influence the capture efficiency of ultrafine particles. In general, ‘4?nines’ (i.e. 99.99%) particle collection efficiency is achievable with baghouse filtration16 and it is considered the best?available technology for particulate capture from the flue gas.15 Twisted Oak is proposing to voluntarily meet a particulate emission standard of 0.03 lb/MMBtu fuel input, which is the most stringent biomass boiler particulate emission standard currently under federal law. A continuous particulate emissions monitor (such as an opacity meter) will also be deployed. Note that current regulations (and commercial particulate monitoring techniques) do not provide any specifications on the total number of particles emitted, or on detailed particle size characteristics, but rather are based on the mass emission rate within certain size classifications (PM10 and PM2.5, for example).

The article by Dorge et al.17 that has apparently received much attention among the citizens of Jasper concerns a thermogravimetric analyzer (TGA) study of Miscanthus pyrolysis and subsequent aerosol formation. This study has almost no relevance to the combustion behavior of Miscanthus in a boiler, because of the low heating rates used in the TGA (20 °C /min). As reported by Dorge et al., the primary aerosol formation in their study occurred over the temperature range of 180 – 360 °C, for which the aerosols that are measured must have been primarily organic in nature (i.e. high molecular weight tar species released during pyrolysis that subsequently condensed upon being cooled). In a boiler, such organic pyrolysis products are consumed through combustion reactions and don’t condense into an aerosol because they are entrained into higher temperature combustion zones, rather than cooled, as occurred in the Dorge study. Even if some of the pyrolysis products did condense into an aerosol in the boiler, the high surface area associated with small particles ensures that they would rapidly be consumed by combustion reactions upon mixing with hot, oxidizing gas. Also, as explained above, the formation of fine particles within a boiler has only an indirect correspondence with the emission of particles from a boiler with particulate control devices.

HCl

HCl emissions are also an area of concern when burning biomass (or when burning any chlorine?containing fuel, including coal). HCl emissions tend to scale with the chlorine content of the fuel, although peak boiler temperature and the alkali and sulfur content of the fuel also play important roles in determining the quantity of HCl emission. At lower temperatures, alkali that is released in the boiler will tend to react with chlorine to produced alkali chlorides, thus removing a portion of the chlorine that would otherwise be emitted as HCl. Brunner et al.12 have performed measurements in a 350 kW moving grate boiler and found 96% Cl capture through the formation of KCl particles. At high temperatures, the alkali preferentially reacts with sulfur to produce sulfates, thereby removing available alkali from reacting with chlorine once the entrained particles are convected to cooler portions of the boiler. In addition, boiler deposits that may initially be composed of KCl are typically sulfated over time,

rereleasing Cl that may be emitted from the boiler as HCl.12,14,15 As discussed previously, Miscanthus has relatively low chlorine content and alkali content in comparison to other herbaceous biomass sources.2?4 Twisted Oak has proposed using dry sorbent injection (using either Na? or Ca?bicarbonate) as a means of capturing chlorine within the boiler and reducing HCl emissions.1 Volatilized chlorine readily reacts with the sorbent and is then captured in the boiler’s particulate capture system. This is the state?of?the?art approach for reducing HCl emissions and is employed in European biomass boilers.15,18 With the high efficiency particulate capture system that is planned for the Jasper boiler, the sorbent injection approach should be quite effective at capturing volatilized chlorine.

NOx

NOx emissions can be significant from any combustion process. For solid fuel combustion, the nitrogen content of the fuel itself contributes to and often dominates NOx formation.18,19 The temperature of the combustion process and the amount of excess air in the main combustion zone are also important factors contributing to NOx formation. The nitrogen content of Miscanthus is typically four times lower than that of other herbaceous fuels and similar to that of wood.2?4 Combustion tests of Miscanthus show moderate NOx emissions, equivalent to that from burning wood.4 The use of overfire air (thereby reducing the stoichiometry in the main combustion zone) helps to reduce NOx emissions, as does flue gas recirculation (FGR) below the grate.15,18 Twisted Oak indicates that they will consider employing FGR, but have not committed to doing this and will await further analysis before determining if and how to employ FGR.11 FGR is typically employed in biomass stoker grate boilers in Europe, both above and below the grate.12,15 Selective non?catalytic reduction (SNCR) is proposed by Twisted Oak to control NOx emissions.1 This is a commonly employed, state?of?the?art approach to NOx emission control.15

Dioxins and furans

Substantial concern has been expressed with regard to the potential for dioxin/furan (PCDD/F) emissions from burning biomass such as Miscanthus in the Jasper boiler. While it is true that Miscanthus has a small chlorine content and thus there is the potential for PCDD/F formation, there is no reason to believe that there will be significant emissions of these highly toxic compounds. In general, incineration of municipal, hazardous, and hospital waste streams are the predominant source of anthropogenic dioxin emissions.20,21 On the other hand, estimates of dioxin/furan emission inventories have consistently identified biomass combustors as significant sources of these compounds.22 A more refined look at the data suggests that most of the dioxin emissions from biomass sources are associated with open?air combustion or with combustion of preservative?treated wood.22 Clean wood burned in well?operating boiler systems with baghouse particle filters typically produces less than 0.1 ng TEQ/dscm (11 vol?% O2 basis), in comparison to waste wood combustion, which can easily yield 100 times higher emissions, even with comparable particulate capture devices present.22

As is well?documented, the formation of PCDD/F requires the existence of four conditions: (a) chloride, (b) aromatic products of incomplete combustion (such as tar compounds), (b) oxygen, and (c) sufficient residence time in a temperature window of approximately 250 – 500 °C.18,21 The requisite residence time for reaction is generally provided by reaction on deposits on surfaces. Therefore, a well?operating boiler,

with well?controlled fuel and air supply (avoiding the periodic formation of fuel?rich ‘puffs’) can avoid formation of significant quantities of PCDD/F’s. The principal difficulties of waste incinerators with formation of these compounds lies in the elevated concentrations of chlorine in these systems and the prevalence of fuel?rich puffs on account of the difficulty of maintaining a consistent fuel feed rate. The monitoring of CO that is proposed by Twisted Oak is a perfectly acceptable means of ensuring that fuel?rich ‘puffs’ are not produced in the boiler, as high concentrations of CO are immediately produced when the local fuel?air stoichiometry becomes fuel?rich. In addition to closely monitoring CO emissions from the boiler, good maintenance of the heat transfer surfaces, preventing the formation of extensive deposits, is good operating practice for both keeping the boiler heat rate low and for minimizing the formation of PCDD/F emissions. With good operational practice, there is no reason to have particular concern over PCDD/F emissions from operating the Jasper boiler on Miscanthus.

Conclusions

The Twisted Oak repowering proposal for the Jasper boiler appears to be well?considered and a promising approach for revitalizing this city asset. Significant challenges exist in the areas of fuel feeding of herbaceous fuels such as Miscanthus into the boiler and in avoiding the formation of excessive slagging and fouling. There is also some justified concern over ultrafine particulate emissions from the boiler when operating on biomass. The proposed air pollution control equipment represents the state?of?the?art in pollution prevention for grate?fired boilers and should adequately protect the citizens of Jasper.

References

1. “Site Lease and Repowering Proposal,” submitted to City of Jasper Municipal Utility Department, Twisted Oak Corporation, Dec. 10, 2010.

2. J.M. O. Scurlock, “Miscanthus: A review of European experience with a novel energy crop,” Oak Ridge National Laboratory internal report, ORNL/TM?13732, Feb. 1999.

3. J. Dahl, I. Obernberger, “Evaluation of the combustion characteristics of four perennial energy crops (arundo donax, cynara cardunculus, miscanthus x giganteus and panicum virgatum),” 2nd World Conference and Exhibition on Biomass for Energy, Industry and Climate Protection, May 10?14, 2004, Rome, Italy.

4. S. Collura, B. Azambre, G. Finqueneisel, T. Zimny, J.V. Weber, “Miscanthus x giganteus straw and pellets as sustainable fuels: Combustion and emission tests,” Environ. Chem. Lett. 4 (2006) 75?78.

5. L. Carvalho, J. Lundgren, E. Wopienka, M. Öhman, “Challenges in small?scale combustion of agricultural biomass fuels,” 9th International Conference on Energy for a Clean Environment, July 2?5, 2007, Póvoa de Varzim, Portugal.

6. E. Wopienka, “Agricultural fuels – hope & reality,” World Sustainable Energy Days, Feb. 25?27, 2009, Wels, Austria.

7. A. Prochnow, M. Heiermann, M. Plöchl, T. Amon, P.J. Hobbs, “Bioenergy from permanent grassland – A review: 2: Combustion,” Biores. Technol. 100 (2009) 4945?4954.

8. E. Heaton, N. Boersma, J. Lok, R. Cruse, J. Singer, “Greener grass? Addressing problems in miscanthus cultivation,” seminar at Univ. of Illinois, April 11, 2011.

9. “A Big Break for Miscanthus Farming,” New Energy and Fuel, Jan. 12, 2011.

10. E. F. Kristensen, L. Fenger, “ Combustion test with Miscanthus in CHP?plant,” 1999, European Energy Crops Internetwork, http://www.eeci.net.

11. Jay Catasein, “6_30 Response to Boiler Questions,” document provided to Bud Hauersperger, June 30, 2011.

12. T. Brunner, F. Biedermann, I. Obernberger, “Combustion characteristics of Miscanthus based on lab?scale and pilot?scale combustion trials in Austria,” presented at the International Energy Agency Bioenergy Agreement, Task 32 Working Group Meeting, May 4?5, 2010, Lyon, France.

13. B.M. Jenkins, L.L. Baxter, T.R. Miles, Jr., T.R. Miles, “Combustion properties of biomass,” Fuel Proc. Technol. 54 (1998) 17?46.

14. L.L. Baxter, T.R. Miles, T.R. Miles, Jr., B.M. Jenkins, T. Milne, D. Dayton, R.W. Bryers, L.L. Oden, “The behavior of inorganic material in biomass?fired power boilers: field and laboratory experiences,” Fuel Proc. Technol. 54 (1998) 47?78.

15. I. Obernberger, J. Fluch, T. Brunner, “Comparative characterization of high temperature aerosols in waste wood fired fixed?bed and fluidized?bed combustion systems,” Proc. 17th European Biomass Conf. & Exhib. Hamburg, Germany, June/July 2009.

16. T. Lind, J. Hokkinen, J.K. Jokiniemi, “Fine particle and trace element emissions from waste combustion – comparison of fluidized bed and grate firing,” Fuel Proc. Technol. 88 (2007) 737?746.

17. S. Dorge, M. Jeguirim, G. Trouvé, “Thermal degradation of Miscanthus pellets: kinetics and aerosols characterization,” Waste Biomass Valor. 2 (2011) 149?155.

18. C. Yin, L.A. Rosendahl, S.K. Kaer, “Grate?firing of biomass for heat and power,” Prog. Energy Combust. Sci. 34 (2008) 725?754.

19. P. Glarborg, A.D. Jensen, J.E. Johnsson, “Fuel nitrogen conversion in solid fuel fired systems,” Prog. Energy Combust. Sci. 29 (2003) 89?113.

20. A.K.D. Liem, J.A. van Zorge, “Dioxins and related compounds: Status and regulatory aspects,” Environ. Sci. & Pollut. Res. 2 (1995) 46?56.

21. J.D. Kilgroe, “Control of dioxin, furan, and mercury emissions from municipal waste combustors,” J. Haz. Mater. 47 (1996) 163?194.

22. E.D. Lavric, A.A. Konnov, J. De Ruyck, “Dioxin levels in wood combustion – a review,” Biomass Bioenergy 26 (2004) 115?145.