Reducing the negative human-health impacts of bioenergy crop emissions through region-specific crop selection

As bioenergy crops continue to replace both existing agricultural crops and natural landscapes, the choice of crop species will become increasingly important given their likely impacts on air quality and climate. The use of crops such as poplar, eucalyptus, and switchgrass as bioenergy feedstocks has increased globally over the past decade, and while ongoing adoption trends are highly dependent on both economic outcomes and policy decisions, significant increases are expected to continue (Energy Information Administration 2013). The large scale land-use changes associated with bioenergy production will have consequences on many aspects of environmental and human health, including food supply, watershed cleanliness, soil quality, and ecological diversity (e.g. Service 2007, Dominguez-Faus et al 2009, Ditomaso et al 2010, Pedroli et al 2013). In some cases, the environmental costs associated with bioenergy production and consumption have been estimated to outweigh the benefits of reduced fossil fuel combustion completely (Hill et al 2009, Tessum et al 2014). In addition to these concerns, many of the most popular candidates for bioenergy crop feedstocks are high-emitters of isoprene, monoterpenes, and other biogenic volatile organic compounds (BVOCs), precursors of surface-level ozone (O₃) and fine particulate matter (PM_2.5). Recent observational and modeling studies of existing bioenergy crop plantations have highlighted the importance of examining ambient chemical and climatological conditions, along with crop emission profiles, when assessing the potential for large-scale cultivation to effect negative air quality consequences in any given area (Hewitt et al 2009, Porter et al 2012). Thus far, however, strategies to inform regional energy crop selection and limit the negative impacts of bioenergy cultivation on air quality and human health have not been adequately addressed.

With a global single-year death toll of seven million, air pollution has now been identified as the most significant environmental risk to human health (WHO 2014). Over half of the total estimated air pollution related deaths are linked to outdoor exposure, particularly to elevated levels of O₃ and PM_2.5 (Brauer et al 2012). These pollutants largely are secondary, forming in situ as a result of chemical reactions involving both naturally and anthropogenically emitted precursors (Sillman 1999, de Gouw and Jimenez 2009), making effective control of their ambient concentrations particularly challenging. For both O₃ and PM_2.5, fluxes of BVOCs can play a crucial role in determining peak daily pollutant concentrations. While all plants emit BVOCs to some extent, there is enormous variability in the magnitude of emission rates between different plant species (Guenther et al 1995). Thus, large-scale landscape-level changes in the type and quantity of plants can have profound consequences on regional air quality by affecting the rates of BVOC emissions, and therefore rates of O₃ and PM_2.5 formation (Wiedinmyer et al 2006, Ashworth et al 2012).

While recent work has examined the air quality consequences of both theoretical and observed land-use changes driven by bioenergy crop cultivation (Porter et al 2012, Ashworth et al 2013), a global study comparing targeted likely bioenergy crop emission profiles has not yet been performed. This comparison is important, since the differences between the highest and lowest emitting candidate crops are large: eucalyptus and other woody energy crops rank among the highest of known BVOC emitters (Guenther et al 1994, Street et al 1997, Padhy and Varshney 2005, Hewitt et al 1990, Winters et al 2009, Owen 2001), while rapidly growing cellulosic alternatives such as switchgrass (Panicum sp.) and Miscanthus (Miscanthus x Giganteus) are among the lowest (Graus et al 2011, Eller et al 2011). Since the conversion of existing fossil-fuel based energy sources to these modern bioenergy cropping systems will presumably be driven by climate and air quality concerns, choosing feedstocks that do not undermine these efforts through drastic changes in biogenic emissions will be critical. As modern bioenergy methods continue to be developed and applied, the regional selection of feedstock crop will naturally be based on a number of agronomic criteria, including growth rates, hardiness, ease of cultivation and harvesting, and suitability for efficient conversion to energy. It is the purpose of this paper to assess whether biogenic emissions must also be added to emerging feedstock crop-selection criteria in areas demonstrating high air quality sensitivities.

In this work, the air quality and climatic impacts of increased biogenic emissions due to a global transition to bioenergy crop cultivation are examined, and considerations for reducing these impacts are proposed and evaluated. To compare these impacts, we model the climate and air quality consequences of large-scale conversion of underutilized land to bioenergy crop cultivation using the NCAR chemistry climate model Community Earth System Model (CESM) 1.1, including version 4 of the Community Atmosphere Model (Lamarque et al 2012, Neale et al 2013; additional model and crop replacement details available in supplementary data available at stacks.iop.org/ERL/10/054004/mmedia). We use three characteristic replacement crop emission profiles representing a range of emission types: trace isoprene emissions (LOW case), high isoprene emissions (ISO case), and high isoprene and monoterpene emissions (ISO + MT case). We also generate a fourth case using an average of all three emission types to represent an intermediate emissions scenario (MIXED). We then use each of these four profiles to simulate varying degrees of land-use change in areas identified as underutilized (Cai et al 2010) under both current and projected future anthropogenic emissions scenarios (Meinshausen et al 2011), and compare them to base cases using non-emitting land rather than a replacement crop.

We run a total of three base cases and 48 replacement cases (table 1), representing a wide range of total replaced area (143–1430 Mha) and emissions profiles (low to high BVOC emitters) using both current and projected anthropogenic emissions and ambient temperatures. Projections of future total biomass energy production potentials are highly variable, mostly due to uncertainties in quantifying land availability and likely crop yields (Berndes et al 2003). For much of this study, the 'Moderate' replacement scale, representing 25% of the full potential area, or around 277 Mha (supplementary table 1 available at stacks.iop.org/ERL/10/054004/mmedia), is examined in detail. Recent literature estimates the area of global abandoned agricultural land at approximately 400 Mha (Campbell et al 2008), making the Moderate case a relatively conservative estimate. We analyze the global model results for six regions with the largest estimates of potentially available land area: Africa, China, Europe, India, South America and the United States. Estimates of net energy gain (NEG) per hectare of energy crop vary widely, even within individual species (Collura et al 2006, Lewandowski and Schmidt 2006, Angelini et al 2009, Schmer et al 2009, 2014, US Department of Energy 2011), but conservative NEG values suggest that land-use conversion in the most aggressive scenarios could generate approximately 10% of current global electricity production.

In addition to comparing pollutant changes between crop profiles and anthropogenic emissions scenarios, we perform an analysis of human-health impacts of alternative cropping scenarios. Using population distribution data and concentration response estimates we estimate the number of lives affected by degraded air quality and assess a health cost on a regional basis (Jerrett et al 2009, Krewski et al 2009, Anenberg et al 2010). We then weight changes in O₃ and PM_2.5 by population density and normalize results by replacement scale to generate an aggregate 'energy crop air quality score' (ECAQS) for each crop and region. Following the examination of urban tree selection and air quality effects presented in Donovan et al 2005, the ECAQS score can be understood as the population-weighted average change in air quality (measured relative to WHO standards for O₃ and PM_2.5) per 10 Mha of energy crop planted in that region, allowing for a normalized measure of AQ sensitivity to variation in crop type and regional abundance. An ECAQS score of negative ten, for example, indicates that the population will see an average increase in pollutant concentrations equivalent to 10% of their respective standards.

3.1. Air quality impacts

Our model results show highly non-linear, region-specific air quality impacts of bioenergy crop selection (figure 1), demonstrating the need for regional crop selections taking local sensitivities into account. Simulated O₃ generally increase for all three high-emitting crop types, with especially large summertime effects predicted in the following regions: India (average increases of 5–27 ppb population-weighted summertime O₃ for the Moderate replacement cases), China (1–9 ppb O₃), and the United States (1–6 ppb O₃). However, changes in O₃ levels are low or even negative in areas exhibiting already low NO_x/volatile organic compound (VOC) ratios, such as South America, where changes in O₃ range from −1 ppb in the summer to +1 ppb in the winter for the Moderate replacement scale ISO and ISO + MT cases, respectively. Secondary organic aerosol formation in all regions is primarily driven by monoterpene emissions, characteristic of the ISO + MT and MIXED cases. ISO replacement cases show some PM_2.5 response, but at much reduced levels, while the LOW cases show negligible impacts on PM_2.5 concentrations. Unlike O₃, PM_2.5 levels consistently increase with greater crop replacement area. PM_2.5 levels also show much less sensitivity to future changes in climate and emissions. In China and India maximum population-weighted changes in PM_2.5 approach 2 μg m⁻³ for the ISO + MT case.

Zoom In Zoom Out Reset image size

Since the health risks associated with PM_2.5 exceed those of tropospheric O₃ relative to the magnitudes of changes evident in this study, the high monoterpene emissions of the ISO + MT cases result in the greatest estimated health costs among the four crop types, with over 410 000 premature deaths worldwide associated with the Moderate replacement level (see supplementary data for more details on mortality estimate calculations (available at stacks.iop.org/ERL/10/054004/mmedia)). Using the reduced emissions of the ISO-only profile leads to 328 000 premature deaths, while conversion of land using LOW emissions shows only very small changes in O₃ and PM_2.5, and therefore negligible increases in premature deaths. Notably, mixing the three crop types (MIXED) proves to be more detrimental than might be expected, with mortality increases almost identical to those of the ISO case worldwide, and greater than the average of the three emissions cases for each individual region. This may be related to nonlinearities in pollutant formation, and therefore in the resulting changes in mortality estimates. The scale of estimated land availability and dense population distributions in India, Africa, and China makes health effects in those areas the most sensitive to increased VOC emissions from bioenergy crops. These areas, already among the most affected by outdoor air pollution (Cohen et al 2005), account for approximately 70% of the increased mortality worldwide.

The extent to which air quality concerns should be taken into account in the selection of bioenergy crops on a regional basis can be evaluated quantitatively by comparing differences in the ECAQS for each crop and region (table 2). Total scores for ISO + MT emissions at the Moderate replacement scale range from −0.9 in South America, where low NO_x levels and a fairly disperse human population keeps health impacts modest; to −14.4 in India, where the density of human population and replacement areas make the air quality consequences of crop selection an especially important factor.

Table 2.  Average Energy Crop Air Quality Scores by region and pollutant for the Moderate replacement scale cases. ECAQS represents change in population-weighted, standard-normalized air quality per 10 Mha of energy crop planted. Peak column represents the maximum monthly average change by crop and region.

	Africa				China				Europe
	O3	PM2.5	Total	Peak	O3	PM2.5	Total	Peak	O3	PM2.5	Total	Peak
ISO + MT	−0.6	−1.0	−1.6	−2.2	−1.8	−2.1	−3.8	−6.1	−0.8	−0.6	−1.4	−3.1
ISO	−0.6	−0.6	−1.2	−1.7	−1.9	−1.1	−3.0	−4.7	−0.8	−0.5	−1.4	−3.0
LOW	0.0	0.0	0.0	0.0	0.0	0.0	0.0	−0.1	0.0	0.0	0.0	0.0
MIX	−0.5	−0.5	−1.0	−1.3	−1.8	−1.2	−3.0	−4.8	−0.6	−0.4	−1.0	−2.3
	India				South America				United States
	O3	PM2.5	Total	Peak	O3	PM2.5	Total	Peak	O3	PM2.5	Total	Peak
ISO + MT	−7.5	−6.9	−14.4	−23.1	−0.1	−0.8	−0.9	−1.2	−1.2	−1.7	−2.9	−5.2
ISO	−7.6	−3.8	−11.4	−18.3	−0.1	−0.6	−0.7	−1.0	−1.3	−1.0	−2.3	−4.2
LOW	0.0	−0.1	−0.1	−0.2	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
MIX	−6.7	−3.6	−10.3	−16.4	−0.1	−0.5	−0.5	−0.8	−1.1	−1.0	−2.1	−3.7

The potential increase in productivity of higher-emitting crops when compared to low-emitting crops such as miscanthus or switchgrass may be justifiable when the population-weighted impacts of the higher emissions are relatively low, as in Africa and South America. On the other hand, the larger impacts shown in the other four regions suggest that finding low-emitting solutions may be crucial for any region-wide implementation of a bioenergy cultivation strategy. Furthermore, comparing the ISO + MT and ISO cases highlights the relative importance of monoterpene and isoprene emissions in each region. While PM_2.5 impacts are larger for the ISO + MT case in all regions, in some areas the health effects of increased O₃ may outweigh those of PM_2.5, reducing the difference between isoprene-only emitters such as giant reed and monoterpene emitters such as eucalyptus. The combined health benefits of using lower-emitting crops for bioenergy production is evident in figure 2, where summed impacts of O₃ and PM_2.5 changes for the ISO + MT case are shown in red, with regional benefits of lower-emitting options (in deaths prevented per energy produced) shown as inset bar plots.

Zoom In Zoom Out Reset image size

3.2. Climatic impacts

3.3. Total costs

Finally, we perform economic assessments of climatic and human-health impacts using estimates for the social cost of carbon, and value of a statistical life. Literature values for both of these costs vary greatly, depending heavily upon future economic and atmospheric assumptions. For this work we use conservative, low-end estimates of $21 per ton CO₂ (IAWG 2010) and $1 million per statistical life (Viscusi and Aldy 2003). We then convert the climatic and human health costs associated with each modeled crop type into equivalent CO₂ (CO₂e) and compare them to estimated benefits of coal combustion reductions (Burnham et al 2012), to produce a comparison of costs and benefits for each crop emission type (figure 3). Costs associated with increased CH₄ (minus the cooling effects of increased organic aerosol) are approximately 8–10% of the expected value of reduced fossil fuel emissions for ISO + MT, ISO, and MIXED emission cases, and negligible for the LOW case. Globally summed health costs are larger, totaling 24–45% of the dollar value of reduced fossil fuel emissions for the high BVOC-emitting cases. These costs make the reduction in negative climate and human health impacts associated with a low-emitting bioenergy crop like switchgrass especially attractive. Under these assumptions, high-emitting crops would dramatically mitigate any net benefits of greenhouse gas emission reductions, primarily through impacts on human health.

Zoom In Zoom Out Reset image size

3.4. Land-use change assumptions and uncertainties

As alternatives to fossil fuels are increasingly adopted, it will be crucial to do so in ways that seek to maximize carbon reduction while also minimizing associated negative consequences. The results of this study suggest that, while cultivation of high productivity bioenergy crops may offer an attractive option for biomass-based energy production in many areas of the world, air quality and climate impacts of crop emissions must also be considered alongside other concerns such as water availability, biodiversity, and food security. In terms of biogenic emissions, using a low-emitting crop such as switchgrass or Miscanthus rather than a higher-emitting alternative could maximize any expected environmental benefits of reduced fossil fuel use. However, the possibility of lower potential yields of these crops compared to higher VOC-emitting options may affect the final cost-benefit analysis of any proposed bioenergy development. By evaluating the local and global consequences of crop selection on a case-by-case basis, including calculation of ECAQS values for candidate crops as performed here, these considerations can help to further inform energy policy and maximize the benefits of alternative energy production while minimizing the negative impacts of bioenergy crop cultivation on air quality, human health, and climate.