Life Cycle Assessment of Biomass-based Combined Heat and Power Plants
Centralized Versus Decentralized Deployment Strategies
Summary
Norway, like many countries, has realized the need to extensively plan its renewable energy future sooner rather than later. Combined heat and power (CHP) through gasification of forest residues is one technology that is expected to aid Norway in achieving a desired doubling of bioenergy production by 2020. To assess the environmental impacts to determine the most suitable CHP size, we performed a unit process-based attributional life cycle assessment (LCA), in which we compared three scales of CHP over ten environmental impact categories—micro (0.1 megawatts electricity [MWe]), small (1 MWe), and medium (50 MWe) scale. The functional units used were 1 megajoule (MJ) of electricity and 1 MJ of district heating delivered to the end user (two functional units), and therefore, the environmental impacts from distribution of electricity and hot water to the consumer were also considered. This study focuses on a regional perspective situated in middle-Norway's Nord- and Sør-Trøndelag counties. Overall, the unit-based environmental impacts between the scales of CHP were quite mixed and within the same magnitude. The results indicated that energy distribution from CHP plant to end user creates from less than 1% to nearly 90% of the total system impacts, depending on impact category and energy product. Also, an optimal small-scale CHP plant may be the best environmental option. The CHP systems had a global warming potential ranging from 2.4 to 2.8 grams of carbon dioxide equivalent per megajoule of thermal (g CO2-eq/MJth) district heating and from 8.8 to 10.5 grams carbon dioxide equivalent per megajoule of electricity (g CO2-eq/MJel) to the end user.
Introduction
Reliance on nonrenewable energies and the effects of global warming have created an unprecedented call of alarm from scientists, political leaders, and activists all around the world. One sector of key significance to our modern society is the electricity- and heat-generating industries; the global electricity mix is dominated by fossil fuels. A key goal toward the avoidance of climate change is therefore incentivizing the use of renewable energies that can replace these fossil sources. Norway, which generates the vast majority of its electricity from hydropower, has also set out ambitious goals to achieve substantial growth in the bioenergy sector (NME 2007). Expansion of biomass to heat and power in Norway is well merited for several reasons: The country engages in significant electricity trading with the rest of Western Europe, domestic heating sources in Norway are dominated by electricity, and the high saturation of large-scale hydropower throughout the latter half of the 20th century has caused policy makers to cope with the environmental objective of preserving the remaining natural riverways. Furthermore, Norway's absolute electricity demand has been increasing from 1% to 1.5% annually.
With abundant forest resources, Norway has an opportunity to develop nonintermittent, renewable baseload heat and power generation from biomass gasification combined heat and power (CHP) plants. Scale is a concern, however, with regard to policy making; on the one hand, although decentralized energy systems are expanding and being promoted at a great rate in many regions of the world (Sims et al. 2007), large-scale as opposed to small-scale CHP solutions are also gaining strong regulatory incentives—for example, with the European Industrial Bioenergy Initiative (EIBI 2011). Quantifiable determination of which scale of CHP plant is most environmentally preferable would provide a good check to ensure that current national policy mechanisms are aligned well with what is best for the environment in terms of energy plant scale. It is therefore important to understand the life cycle environmental performance of CHP gasification systems and whether the scale of CHP plays a considerable role.
A recent review article (Cherubini and Strømman 2011) shows that past LCA research involving heat and power production from biomass is well represented in the literature. A considerable number of studies within the last decade have focused on electricity production from dedicated short-rotation bioenergy crops (Uihlein et al. 2008; Goglio and Owende 2009). For instance, Uihlein and colleagues (2008) found that there was no clear environmental advantage between dedicated bioenergy crops (corn and miscanthus) and conventional fossil fuel for several energy-related products, although in terms of mitigating climate change, generally some advantages were foreseen. Zhang and colleagues (2010) integrated life cycle assessment (LCA) with life cycle costing (LCC) to compare several coal and wood pellet cofiring mixes with 100% pellet-based power production; they found that carbon dioxide (CO2) mitigation was both substantial and competitive with other renewable options and that both the infrastructure and the regional pellet production were in place for near-term implementation. These studies vary in impact and LCA methodology, but all consider a unit process-based approach. Impact categories from the CML 2000 baseline (Hischier et al. 2009) methodology are prevalent, but only a few widely-used categories were considered. The most emphasis in terms of results is placed on the energy balance and net CO2 equivalent (CO2-eq) emissions. Several recent studies have also produced LCA results for stand-alone biomass heating systems (Eriksson et al. 2007; Solli et al. 2009; Caserini et al. 2010). Solli and colleagues (2009) carried out a hybrid LCA on wood stoves for the region of Norway, and Eriksson and colleagues (2007) and Caserini and colleagues (2010) carried out process-based attributional LCA on some district heating options.
Another common feature of the scientific literature on LCA of bioenergy systems is the assumption that CO2 emissions are carbon neutral, and therefore, do not contribute to climate change, which means that biogenic CO2 is not accounted for as a greenhouse gas (GHG) emission. This is done in keeping with the basic principle that CO2 released from combustion will be removed from the atmosphere by biomass regrowth. Despite some recent criticisms of this accounting procedure (Möllersten and Grönkvist 2007; Cherubini et al. 2011), in this article we follow the mainstream practice in LCA of bioenergy systems and consider biogenic CO2 emissions as climate neutral because we assume that regrowth occurs. Given that Norway has been witnessing a significant forest stock expansion, with annual incremental additions increasing at a rate of 1.3% per year over the past 50 years (SSB 2007), the carbon pool neutrality assumption is safe for this study.
Optimization of bioenergy scale has also been dealt with thoroughly from a cost perspective by Searcy and Flynn (2009); their work indicates that there is indeed a clear trade-off between economy of scale of the energy plant and the biomass procurement costs due to increasing transportation distances. Similar trade-offs between scales of deployment may also show important differences in terms of life cycle environmental impact. Also, in terms of energy distribution from plant to end user, Persson and colleagues (2006) found significant life cycle environmental impacts associated with operational energy distribution losses and infrastructure. Although small-scale, as opposed to large-scale, energy plants would have a lower operational efficiency, smaller scale energy plants could require a shorter average distance from energy plant to end user and therefore may contribute fewer energy distribution impacts than larger scale plants.
It is therefore the objective of this study to answer the following question: What are the comparative life cycle impacts between several scales of biomass gasification CHP plants, with energy delivered to the doorstep of the end user? Building on the state of the field, this study incorporates the grid and district heating capital and the difference between CHP plant scales into the analysis of biomass CHP. That is, a unit-process based LCA was implemented and compared across micro (less than 1 megawatt [MW]1), small (1 to 20 MW), and medium scales (20 to 100 MW) and was loosely connected to the geographical context of a middle-Norway region (the counties of Nord- and Sør-Trøndelag). The methodology, along with key system assumptions, is presented in the following section, and the results and analysis are illustrated in the next section. The article concludes with a discussion of the results.
Methodology and Case Description
LCA is a prevailing tool in analysis of renewable energy systems for environmental impacts; its principles, framework, and requirements are defined as explained by Guinée and colleagues (2002). In this study, we use attributional LCA to analyze the environmental impacts of the hypothetical case of one micro, one small, and one medium scale CHP plant fuelled by forest residues in the mid-Norway region. Attributional LCA was deemed to be the best approach because this is a unit-based study; in addition, because the main goal of this research was to compare several scales of CHP, a detailed description of the stand-alone systems was suitable. CHP gasification technology was chosen because it is considered the best available bio-CHP technology for micro to medium scales due to high power-to-heating ratios and overall efficiencies. For the micro and small-scale systems, down-draft gasification was chosen, and for the medium scale, integrated gasification combined cycle technology was assumed (see the work of Kwant and Knoef [2004] for the European Union [EU] and Knoef [2005] for a technology description and global experience on gasification projects applied at the micro to medium scale).
The objective of this study was to do a unit-based analysis of the life cycle environmental impacts due to 1 MJ of thermal energy (MJth) and 1 MJ of electricity (MJel) delivered to the end user. Figure 1 illustrates a simplified flowchart of the system. The major processes that have the most significant differences among the three scales include average biomass procurement distance (longer distance for larger scale), grid and district heating losses (higher losses for larger scale), CHP plant emissions (higher rate per energy unit for smaller scale), average electricity and district heating distribution distance (longer distances for larger scale), and the overall electrical and thermal heat efficiencies of the CHP plants (lower efficiency for smaller scale; see figure 2). Unit process-based factors stemming from empirical and simulation engineering data are assumed. The ecoinvent v2.0 database (Ecoinvent 2010) is used for the majority of the foreground processes (see the Supporting Information on the journal's Web site). All processes considered to be occurring locally (within Norway) have been modified under the assumption that Norwegian-produced electricity and fossil oil are used as inputs (Dones et al. 2007).
Foreground system flow diagram.
Main trade-offs to consider among the three scales of combined heat and power (CHP) systems.
Life Cycle Assessment Methodology
The midpoint CML 2 Baseline 2000 impact assessment method (CML 2001) is used to assess these impacts. This method was chosen for its holistic set of impact categories. The categories include global warming potential (GWP), acidification potential (ADP), abiotic depletion potential (ABP), eutrophication potential (ETP), photochemical oxidation potential (PCOP), ozone layer depletion potential (ODP), human toxicity potential (HTP), fresh water aquatic toxicity potential (FWTP), marine aquatic toxicity potential (MTP), and terrestrial toxicity potential (TTP).
Biomass Procurement Distance and Processes
The unit process-based data required for this system begins in the forest, where the residue was assumed to be compressed and tied into bundles, forwarded to the roadside, and then transported by lorry to the CHPs. A Scandinavian softwood forestry process was assumed (Werner et al. 2007), and a modified agricultural baling process was used for forest residue bundling (Nemecek and Kägi 2007). Biomass forwarding was assumed to be carried out with a 3.5 to 16 tonne (t)2 lorry, with operational emissions adjusted to Norwegian industry assumptions (Spielmann et al. 2007; Michelsen et al. 2008). Two sources of fuel were assumed: forest residues (FR) from thinning and harvesting, and sawmill residues (SR) from the local sawmills of the region. For the medium (90% FR and 10% SR), small (70% FR and 30% SR), and micro-scale (70% FR and 30% SR) CHP, the average procurement distances of acquiring FR from forest were then determined to be 115 kilometers (km),3 39 km, and 29 km, respectively. Actual roadway distances from the six largest sawmills in the region to the city of Trondheim, where the medium-scale CHP was assumed to be located, were used for the assumed SR procurement distance, which was calculated according to a weighted average based on the production rate of each sawmill. This average distance was calculated to be 110 kilometers (km). For the small and micro-scale CHPs, the average procurement distance of acquiring SR was assumed to be 15 km. The biomass road transport was assumed with a modified lorry (heavier than 16 tonnes, fleet average [Spielmann et al. 2007]).
Combined Heat and Power Plants
For the micro and small-scale CHPs, a fixed-bed down-draft gasifier was considered, because only dry gas cleaning would be needed. This significantly reduces the wastewater stream at the plant. One such technology has been demonstrated by a Danish company, BioSynergi Proces ApS, and technical data from it are utilized for the micro and small-scale CHPs (Biosynergi 2010). An appropriately sized internal combustion engine integrated with a generator was then considered for the power production. Table 1 below shows the efficiencies, capacity factor, and total energy production assumed for each scale of CHP.
| CHP scale | MWel_cap | MWth_cap | ηel | ηth | ηtot | Capacity factor | Annual energy production (GWh) | Main source |
|---|---|---|---|---|---|---|---|---|
| Micro | 0.1 | 0.22 | 0.24 | 0.52 | 0.76 | 0.51 | 1.40 | Biosynergi, Öko-Institut |
| Small | 1.0 | 2.31 | 0.30 | 0.56 | 0.86 | 0.51 | 15 | Biosynergi, Öko-Institut |
| Medium | 50 | 66.7 | 0.386 | 0.51 | 0.90 | 0.51 | 520 | Öko-Institut |
- Note: MWel_cap, = electrical capacity (in MW); MWth_cap= thermal capacity (in MW); ηel= electrical efficiency; ηth= thermal efficiency; ηtot= overall efficiency.
As CHP scale increases beyond a few megawatts of electrical capacity, more sophisticated and integrated technology has been deemed the most favorable technological option. Already, several integrated gasification combined cycle (IGCC) plants have been built (ranging from 5 to 10 MWel) that use circulating fluidized bed (CFB) gasifiers, and many researchers think that the IGCC concept has the highest potential of all the possible gasification applications (Kwant and Knoef 2004). A pressurized fluid bed with an air stream to gasifier, with integrated gas and steam turbines, is assumed. Because no such plant currently exists at the scale of 50 MWel, efficiency data were based on simulation (Öko-Institut 2010). All operational emissions data for the three CHPs stemmed from emissions inventory of similar gasification CHP processes and scales in the GEMIS 4.5 database (see supporting information on the Web for more detail). All biogenic CO2 emissions were assigned a GWP of zero. The capacity factor of 0.51 for all CHPs was chosen due to seasonal fluctuations in heat demand and is based on actual data from a medium-scale biomass CHP located at similar latitudes in Lugnvik, Sweden (Jämtkraft AB 2008). For all systems, the CHP capital was based on a small-scale CHP plant (wood chips, burned in cogen 6,400 kilowatts thermal [kWth], emission control under Swiss conditions; Jungbluth et al. 2007), and an economy of scaling factor of 0.7 was used (Flynn 2006).
Allocating Environmental Impacts to Combined Heat and Power Products
The purpose of allocation is to distribute the environmental load to the products—in this case, heat and power—in a representative way, because these products are outputs of the same process. Economic allocation, or perceived value of electricity and district heating, was considered, but the price differences can vary to a high degree from region to region due to the naturally monopolistic behavior of district heating (Westin and Lagergren 2002). For example, Trondheim Energi AS regulates district heating at a price that is generally around 10% lower than the electricity price (Trondheim Energi 2010), whereas Jämtkraft AB provides district heating to the majority of Östersund residents in Sweden at a price that is around 30% less than local electricity prices (Jämtkraft 2010). Due to the localized and often vertically integrated nature of the district heating industry, economically based allocation was rejected. Allocation based on energy output was considered, but such a measure does not take into account the quality of the energy, and, therefore, on a per-unit energy product basis equivalent impacts would occur for both district heating and electricity at the plant. Therefore, exergy allocation was chosen for this study (see Ayres et al. 1996). The exergy-based fraction of environmental impacts attributed to electricity, Eel, for the medium, small, and micro scale was calculated to be 0.78, 0.72, and 0.69, respectively. This means, for example, that 22% of the environmental impacts are associated with the medium-scale district heating product, and 78% are associated with the electricity product per useful energy unit produced at the CHP plant (the supporting information on the Web provides the exergy calculations).
District Heating Systems
An extensive unit process-based LCA on district heating systems was utilized in this study (Fröling et al. 2004; Fröling and Svanstrom 2005; Persson et al. 2006). Three sizes of pipes with differing inner diameters (ids) were considered: DN 25 Twin (id = 31.4 millimeter [mm]),5 DN 100 Single (id = 110.4 mm), and DN 500 Single (id = 501.7 mm). (Twin meaning two pipes connected and single meaning one). The assumptions regarding distribution of each size of pipe and the corresponding lengths for each of the pipe sizes was based on overall statistics from Trondheim Energi's district heating operation in 2009 (Trondheim Energi 2010). With 548 gigawatt-hours (GWh)5 of thermal district heating produced in Trondheim's DH system running through 167 km of piping, the length per energy factor assumed for all scales of CHP was 0.305 km/GWh-yr; this value is comparable to the Swedish District Heating Association's nation-wide factor of 0.299 km/GWh-yr (Svensk Fjärrvärme 2009). Distribution of length of pipe sizes was assumed to be the following: Medium scale (25% DN 25 Twin; 62% DN 100 Single; and 13% DN 500 Single); Small scale (40% DN 25 Twin; and 60% DN 100 Single); and Micro scale (100% DN 25 Twin). This distribution was based on the actual piping length and pipe size distribution of Trondheim Energi. District heating losses were based on Fröling and Svanstrom (2005), and network losses for each system were calculated to be 8.1% (medium), 7.0% (small), and 5.0% (micro). The background material and energy processes needed to produce the district heating capital came from the ecoinvent v2.0 database.
Electricity Grid System
Electricity losses differ between the scales of CHP plants because it was assumed that a smaller plant requires a shorter average distance to deliver its electricity to the user. Although it is virtually impossible to actually follow the path of electricity from generator to end user, it was assumed that 100% of electricity production from the CHP is consumed or lost locally, and the area of this locality depends on the annual consumption of the given locale. With estimated lengths and network branches, along with assumed line materials and cross-sections, we used common design equations (Solvang 2008) for electrical losses to estimate the electricity dissipated throughout each of the pathways. For medium, small, and micro-scale CHP, electricity lost through the electrical lines to the end user was then assumed to be 9.3%, 2.4%, and 0.68%, respectively. Electricity mixes and grid infrastructure inventory came from the work of Dones and colleagues (2007; see the supporting information on the Web for further elaboration).
Results
Figures 3 and 4 illustrate a contribution analysis for the three CHP systems considered for both functional units—1 MJ of electricity (figure 3) and 1 MJ of district heating (figure 4) delivered to the end user. The x-axis represents the fraction of environmental impact as normalized by the CHP scale that is contributing the most to the given impact category.
Contribution analysis of the environmental impacts from producing 1 megajoule (MJ) of electricity (top bar [MIC]= micro scale; center bar [SML]= small scale; bottom bar [MED]= medium scale).
Contribution analysis of the environmental impacts from producing 1 megajoule (MJ) of district space heating (top bar [MIC]= micro scale; center bar [SML]= small scale; bottom bar [MED]= medium scale).
For the ten impact categories analyzed in this study, the overall picture between the scales is mixed, and all absolute values are of similar magnitude. Therefore, on a per-megajoule basis, the results for electricity and district heating to end user are quite comparable, but with some important differences. Beginning with the impact category GWP, the analysis of the results is discussed for each impact category.
Global Warming Potential
GWP is dominated by the operational air emissions of the CHP plant: electricity product—49% (medium), 63% (small), 63% (micro); heating product—37% (medium), 50% (small), 60% (micro). The second largest contributing process for the electricity product is biomass transportation (30% for medium, 7.9% for small, and 7.2% for micro), whereas for the district heating product, it is the district heating capital (25% for medium, 21% for small, and 5.4% for micro). Carbon dioxide from fossil sources throughout the system and dinitrogen oxide (N2O) from CHP emissions are the stressors that contribute the most to overall GWP (CO2: 34% to 47% for electricity, and 37% to 58% for heating product; N2O: 44% to 54% for electricity, and 33% to 44% for heating product). For the electricity product, the small and medium scales contribute 84% and 98%, respectively, of the overall impact that the micro-scale system creates. For the district heating product, the small and micro-scale systems contribute 86% and 87%, respectively, of the overall impact that the medium-scale system creates. Most GWP stems from two scale-sensitive processes (biomass transportation and CHP emissions), but, as explained earlier, they are considered trade-offs, and for both energy products the small CHP has the lowest GWP, which suggests an optimized plant size.
Photochemical Oxidation Potential
More than 71% to 93% of the PCOP derives from air emissions from the CHP plant, largely due to carbon monoxide emissions. The biomass procurement chain—from forestry to chipping—contributes 5.0% to 16% to this impact. With regard to the electricity product, medium and small-scale CHP contribute 39% and 60%, respectively, relative to the PCOP impact contributed by micro scale CHP, whereas for the district heating product, the contribution is 41% and 63%, respectively. The results for PCOP show that the smaller the CHP plant is, the higher the impact it has, largely due to less pollution control of CHP air emissions.
Ozone Layer Depletion Potential
Nearly the entire ozone depletion impact (∼96%) comes from two stressors—bromtrifluoro-methane (halon 1301) and bromochlorodifluoro-methane (halon 1211). These chemicals are used as a gaseous fire-suppression agent, and, therefore, this chemical was mostly being used in the upstream crude oil (74% to 95%) and natural gas (1.7% to 19%) procurement processes. For the electricity product, the small and micro scale contribute 61% and 78%, respectively, of the overall impact that the medium-scale system creates. For the district heating product, the micro and small scale contribute 38% and 78%, respectively, of the overall impact that the medium-scale system creates. The impacts that create OZP have been shown to increase with larger CHP scale, due to higher fossil oil and natural gas demands, which largely stem from longer biomass procurement distances, assumed to be energized with fossil diesel fuel.
Eutrophication and Acidification Potential
Both ETP and ADP are dominated by the operational air emissions of the CHP plant: electricity product—52% and 56% (medium), 84% and 84% (small), 92% and 92% (micro); heating product—47% and 43% (medium), 82% and 81% (small), 92% and 92% (micro), respectively. Nitrogen oxides contribute to nearly all of this environmental impact: 89% to 97% for electricity product, and 80% to 97% for heating product. For the electricity product, the medium and small scale contribute 20% and 43%, respectively, of the overall impact that the micro-scale system creates. For the district heating product, the medium and small scale contribute 22% and 45%, respectively, of the overall environmental impact that the micro-scale system creates. Both ETP and ADP have been shown to increase with decreasing CHP scale; this is largely due to higher NOx emissions at the smallest scale plant due to less pollution control technology.
Toxicity Potential
The electricity grid contributes 63% to 87% of the total TTP impact; this is mostly attributed to Chromium VI, which was assumed to make up the preservative used on the surface of the electricity poles. The most problematic process creating MTP and FWTP (around 50% of total impact) was found to be bioash disposal to the sanitary landfill, and heavy metal ions, such as nickel and vanadium, were the major stressors creating such impact. For HTP, the electricity grid dominates this category (12% to 30%); arsenic and polycyclic aromatic hydrocarbons are the constituents of most concern. In terms of sensitivity to scale, the toxicity potentials differ greatly. For TTP the electric grid contributes significantly, and therefore the smaller the scale is the fewer potential impacts there are. For MTP and FWTP, however, the impacts are largely due to bioash disposal, and therefore the larger scale CHP would accrue less impact than a smaller scale CHP system, due to the assumption of more efficient gasification reactors with the larger scale systems. See the supporting information on the Web for more details on each of the toxicity impact categories.
Abiotic Depletion Potential
For the electricity product, biomass transport contributes the largest impact: 62% (medium), 23% (small), and 21% (micro). Conversely, for the district heating product, the district heating capital contributes the largest impact: 49% (medium), 23% (small), and 21% (micro). The significant drop between medium/small and micro scales is due to the distribution of pipe sizes for the district heating network that was assumed. The main abiotic resources (which account for nearly 100% of depleted resources) that are being depleted are fossil fuels: oil (66% to 77%), natural gas (11% to 16%), and coal (6.7% to 16%). For the electricity product, the small and micro scale contribute 59% and 72%, respectively, of the overall impact that the medium-scale system creates. For the district heating product, the micro and small scale contribute 48% and 68%, respectively, of the overall impact that the medium-scale system creates. Like OZP, ABD is highest in whichever system consumes the most fossil fuels. Because the medium-scale system consumes more diesel oil for biomass transportation, the impacts for ADB are higher for the larger scale system.
Direct and Indirect Impacts Associated With Heat and Power Distribution
The impacts from the energy distribution capital are indeed significant, and they can be segregated into direct and indirect impacts: “Direct impacts” means impacts from the manufacturing and construction of the energy distribution capital (i.e., district heating network and electricity grid infrastructure) and those indirectly attributed to the energy distribution due to operational energy losses that create an increased demand on the rest of the system upstream. These impacts can be seen in figure 5, and we calculated average values to illustrate the general significance (i.e., average percentage of total environmental impact across all categories due to indirect impacts): electricity product—7.4% (medium), 2.1% (small), 0.62% (micro); district heating product—5.6% (medium), 5.1% (small), 4.4% (micro). With regard to the total system environmental impacts (i.e., direct from capital and indirect due to energy distribution losses) stemming from the energy distribution networks, the impacts are considerably more apparent (average contribution across all impact categories): due to electricity product—medium scale (21%), small scale (15%), and micro scale (8.8%); due to district heating product—medium scale (31%), small scale (27%), and micro scale (12%).
Environmental impacts due to energy distribution capital and losses from combined heat and power (CHP) plant to end user for electricity production (top) and district heating production (bottom). Percentages above bars represent the percentage of total system impacts due to energy distribution.
Discussion and Conclusion
The objective of this study was to understand the environmental performance of producing 1 MJ of electricity and 1 MJ of district heating delivered to the end user from three scales of biomass-based CHP plants using gasification technology. We showed that there are several trade-offs to consider that make one scale more advantageous in some impact categories but worse in others. We also showed that energy distribution from CHP plant to end user contributes significantly to the overall impacts of the entire system.
In terms of quality of system assumptions, forest residue was the biomass source utilized, but the assumed mix (FR and SR) may significantly change the results, because the impact of producing an energy unit from FR was found to be over four times higher than from a unit of SR. The results also clearly show that impact (namely ODP, ABD, GWP, HTP, and MTP) is highly sensitive to biomass procurement distance. Here we assumed an ideal logistical situation in which the residue supply lay in close proximity to CHP plants. This can and has been shown to be realistic in practice, and it is important that future-plant managers plan biomass procurement strategies with localized sources of biomass as a high priority.
We have illustrated that CHP plant operation contributes significantly to several important environmental impact categories (GWP, ADP, ETP, and PCOP). Original data, as opposed to emissions data from operational gasification CHP plants, are therefore desirable. Impact results were found to be sensitive to conversion efficiencies and capacity factors. Such efficiencies may improve in the short term, although capacity factors could change depending on both climate and thermal end-use application (i.e., higher capacity factor in both warmer climates and more intense year-round industry). Electricity is difficult to model, and therefore several perspectives on how to allocate electricity are held today. More advanced electricity modeling that could more accurately determine the actual average pathway and distance from CHP plant to end user would likely improve the assessment of the environmental performance of the system. Additionally, life cycle inventory (LCI) data for grid-specific inputs in Norway would be optimal. The top contributors to uncertainty were therefore reasoned to be due to operational CHP emissions, biomass procurement distance, and energy distribution distance.
Benchmarking to Global Warming Potential
For electricity production to end user, we found 8.8 to 10.5 g CO2-eq/MJel GWP at the CHP plant; for district heating to end user, we found 2.4 to 2.8 g CO2-eq/MJth GWP at the CHP plant. The results are compatible to those of Dones and colleagues (2007) for a small-scale steam Rankine-cycle biomass fuelled CHP and with exergy allocation: 1 MJ electricity at plant caused 8.4 g CO2-eq/MJel, and 1 MJ district heating at plant resulted in 2.6 g CO2-eq/MJel. Additionally, an extensive review (53 studies) of biopower LCAs (CHP included) undertaken by NREL (2010) found a span of life cycle GWP values for forest residue electricity production ranging from 1.4 to 25 g CO2-eq/MJeq, with a median of around 10 g CO2-eq/MJel. In terms of acidification potential, Pehnt and colleagues (2006) and Dones and colleagues (2008; who used the same system as above) found 65.8 and 75.6 mg SO2-eq per MJel at the plant, respectively, whereas the values in this study ranged from 55 (medium) to 270 (micro) mg SO2-eq per MJel at the plant. Additionally, the difference in GWP impact due to district heating capital was, on average, 120% higher than the impacts calculated by Persson and colleagues (2006). This is likely due to truncation error, given that an extensive LCI database, such as Ecoinvent v.2.1, was not employed in the analysis. The results calculated in this study are therefore within the range from previous studies on bio-heat and power from forest residues.
With regard to scale, important localized impact categories in this system include ADP, ETP, FWTP, MTP, and PCOP, because the majority of environmental impact stems from a single point source (i.e., stressors from air and ash directly from the CHP plant), whereas the important globalized impacts are GWP, ABP, and OZP. Because the smallest CHP plant emitted more emissions/ashes per energy unit, it caused more impact in these localized categories. Conversely, the larger scale CHP plant caused more impact in the globalized categories, and this was largely attributed to longer biomass procurement distance. Therefore, if biomass is locally abundant and sufficient to meet the energy requirements of a given CHP system, the micro to small-scale systems will likely have lower potential globalized impacts but will still be faced with higher localized impact potential.
As illustrated in figure 5, energy distribution from the CHP plant to the end user showed a significant share of the total system environmental impact: from less than 1% (PCOP, micro, electricity) to around 89% (TTP, medium, electricity) depending on the scale, product, and impact category. Apart from ADP and ETP for the district heating product (due to closer energy losses between the scales), the energy distribution of the micro-scale plant created the least impact for all of the other impact categories across both products. Likewise, energy distribution of the micro-scale system created the lowest total system environmental impacts relative to the other scales. Energy delivery to the end user is therefore a very important process in the life cycle of the system and should be included within the system boundaries of an electricity or district heating LCA. This guarantees that the main impacts in the whole system are accounted for while also ensuring a consistent comparison between the different scales of energy system.
Implementation—Barriers and Opportunities
Bioenergy for stationary applications needs considerable help from Norwegian renewable energy policy mechanisms. Low electricity prices and few central heating facilities (around 70% of household heating stems from electricity) are the main reasons why bioenergy in Norway has a significantly lower share of the energy market than in neighboring countries, such as Sweden and Finland (Trømborg et al. 2007). With respect to biomass CHP systems, Trømborg and colleagues (2007) found that CHP investments based on forest fuels would require a 50% reduction in investment costs and nearly a doubling increase in electricity prices to be profitable with the current biomass prices in Norway. Key Norwegian funding institutes that promote bioenergy projects with subsidy programs, namely Enova and Innovation Norway, were found to harmonize well with the results of our study and the scales of bioenergy projects that they promote, offering up to 20 and 35% of investment costs, respectively. Such subsidy ceilings may not be sufficient for small-scale CHP plants at this time, however, whereas strictly bioheating applications will likely benefit most from these programs.
Conclusion
Our analysis shows that there is no rationale for discriminating between scales of stationary bioenergy plants in terms of environmental performance; the associated impact due to energy distribution was decisive in bringing us to this conclusion. This study has shown that on a per-energy-unit basis, the impacts across all scales are well within the same magnitude; larger scale CHPs are higher in the globalized impact categories, and the smaller scale CHPs are higher in many of the localized impact categories. The impact due to energy distribution to the end user was best for the CHP of least scale, and, overall, the associated impacts were very significant for several of the impact categories. On the basis of our analysis, we can conclude that policy mechanisms should be targeted equally at all scales of heat and power plant. In addition, on an environmental level, energy distribution operation and capital should not be excluded from a district heating or power product based LCA, especially when scale is being compared.
Acknowledgements
This study was financed by a PhD stipend from the Norwegian University of Science and Technology (NTNU). We are very grateful for the available funding that made this research possible through the CENBIO program and from the Norwegian Research Council.
Notes
References
About the Authors
Geoffrey Guest is a PhD candidate at the Norwegian University of Science and Technology (NTNU) in Trondheim, Norway. Ryan M. Bright is a PhD candidate at NTNU. Ottar Michelsen is a researcher at NTNU. Francesco Cherubini is a postdoctoral researcher at NTNU. Anders H. Strømman is an associate professor at NTNU. All authors are part of the Industrial Ecology Program of NTNU.
