Air quality and climate benefits of long-distance electricity transmission in China

Most countries now recognize that climate change is a serious problem, and are experimenting with ways to cut greenhouse gas (GHG) emissions that align with their national interests. For most developing countries that means finding ways in which reducing GHG emissions facilitates the achievement of local goals such as reduced air pollution with associated improvements in public health [1]. Given the short time horizon of politicians and the fact that global benefits from reductions in climate change are diffused over all countries, tangible national co-benefits have proved to be more effective in motivating immediate carbon mitigation actions than the long-term threat from climate change [2].

One central finding from the latest report of the Intergovernmental Panel on Climate Change (IPCC) is that electrification will likely play a central role in decarbonizing energy systems [3]. Decarbonization by electrification is possible in many ways, and many governments are looking at much larger roles for renewables—an option that requires careful long-distance transmission planning since renewable resources are often remote from centers of power demand and often variable and intermittent in supply (especially wind and solar power) [4, 5]. They hence often require transmission systems sized much larger than the useful power they consistently deliver, and require local suppliers of electricity to manage reliability—often by using fossil-based power plants that contribute to local air pollution. However, most transmission lines are powered entirely by fossil-based energy at present. Although electricity transmission could help relocate fossil-based generation and associated air pollution to remote areas, fossil fuel powered 'energy by wire' approaches still result in large carbon emissions and global climate impacts. Thus 'energy by wire' strategies may involve tradeoffs between local and global benefits. Our paper develops methods to quantify these tradeoffs.

We focus on China, where national policy has emphasized the need to achieve global carbon mitigation goals in ways that maximize national air pollution benefits [6, 7]. China thus serves as a valuable test case for other emerging economies—such as India, South Africa and Indonesia—that also have large potentials for curbing their growing CO₂ emissions but are wary of adopting expensive policies unless they yield large, tangible local benefits. Moreover, China is a pivotal country in global cooperation on energy and climate. The country is the world's top carbon emitter and largest investor in renewable energy. It has recently made commitments to peak its CO₂ emissions by 2030 [8]. Most importantly, as a country facing severe air pollution, concerns about air pollution have already inspired the government to move coal power plants out of polluted urban areas in the east and to expand investment in long-distance transmission capacity. Twelve 'Electricity Transmission Corridors for Air Pollution Control' are under construction to replace eastern coal power generation with imported electricity in order to reduce local air pollution [9] (figure 1(a) and supplementary table 1, available at stacks.iop.org/ERL/12/064012/mmedia), making it the first time in the world that transmission planning is used for air pollution control. The experience with these transmission systems is likely to guide subsequent investment in China and other countries.

Zoom In Zoom Out Reset image size

Most previous studies on electricity transmission either focus entirely on the carbon emissions [10, 11] or simply tabulate the reduction of local air pollutant emissions without evaluating the impact of those emissions on air quality and human health [12]. Here we not only account for the carbon emissions, but also conduct a sophisticated evaluation of the air quality impacts on health. We use a regional state-of-the-science air pollution model [13] to fully consider the effect of meteorological factors and atmospheric chemistry, and recent epidemiological evidence that finds decreasing health impacts from a unit increase in PM_2.5 (fine particulate matter of 2.5 μm diameter or smaller) at the elevated PM_2.5 concentrations found in highly polluted parts of China (i.e. concave shape of mortality risk functions) [14, 15]. Furthermore, most prior literature has looked at highly stylized scenarios—such as fully coal-by-wire [16] (which reduces local air pollution but has little global benefit) or almost fully renewables such as wind-by-wire [17] (which can be quite expensive given the variability of wind resources). In contrast, our approach examines scenarios that align with current government plans for the near future of the Chinese power system. Based on current plans, nine of the twelve proposed lines will transmit only coal power. Two are planned to transmit wind power along with coal power (though the exact power mix is not specified in the plan), and a third in southern China will be powered entirely by hydroelectricity. We therefore look at a fully coal by wire scenario, and compare it to a hybrid of renewables and coal by wire scenario considering the renewable resource potentials in the power exporting regions. The current government plan is in between these two 'by-wire' scenarios, and likely closer to the fully coal by wire scenario. Due to the geographic mismatch between resource-abundant western provinces and the demand centers in the east, long-distance transmission is critical in both scenarios in order for the eastern provinces to access coal or renewable resources in inland regions. We also compare these two 'by-wire' strategies to 'coal by rail', the most commonly used energy transfer option at present in which coal (abundant in the west of the country) is transported as primary fuel to wealthier, populous centers in the east where it is burned to generate power [18].

2.1. Energy and emission scenarios

2.2. Regional atmospheric chemistry simulation and evaluation

2.3. Evaluating health impacts from air pollution exposure

We focus on ambient fine particulate matter (PM_2.5), the air pollutant with the largest impact on human health [32] (see online supplementary figures 4 and 8 for results on ozone). For four diseases associated with long-term exposure to PM_2.5 (ischemic heart disease, stroke, chronic obstructive pulmonary disease and lung cancer), we use the following equation to calculate the mortality changes relative to the BASE scenario in each province:

$$\begin{equation} \Delta \,{\rm Mortality}_d = I_{d,{\rm BASE}} \cdot {\rm PoP} \cdot \left( {\frac{{{\rm RR}_d (C_s )}}{{{\rm RR}_d (C_{{\rm BASE}} )}} - 1} \right) \end{equation}$$

Pop is the total adult population aged 25 and above in each province, based on county-level China census data [33]. I_d,BASE is the disease-specific (d) annual average national BASE mortality rate in the total adult population. RR_d(C_s) and RR_d(C_BASE) are the relative risks (RR) of disease d for adult population at the PM_2.5 levels of C_s in scenario s and C_BASE in BASE, respectively. I_d,BASE and RR are based on the Global Burden of Disease (GBD) database [15]. We use the four-month average of the simulated population-weighted, provincial-averaged PM_2.5 concentrations to estimate annual mean exposures for each province. Sensitivity analyses on linear RR functions and spatial resolution for exposure assessment are presented in online supplementary figure 11.

3.1. Health-related air quality benefits

3.1.1. Air pollutant emissions

Power plants emit little PM_2.5 directly. We hence present results on precursor pollutants of SO₂ and NO_x, that form secondary PM_2.5 in the atmosphere. In CbR, air pollutant emissions only decline in the eastern importing provinces (6% and 11% reduction in SO₂ and NO_x emissions). In the energy-by-wire scenarios, eastern emissions decline more (15% and 16% reduction in SO₂ and NO_x emissions), while emissions increase in the exporting regions due to added coal power generation (figure 1(b) (SO₂) and online supplementary figure 5 (NO_x)). As renewable electricity generation has no emissions, Hybrid yields smaller increases in power exporting regions (4% and 3% increase in SO₂ and NO_x emissions) than CbW (10% and 8% increase in SO₂ and NO_x emissions). Total emission changes across all importing and exporting regions for Hybrid are a decrease of 6% and 9% for SO₂ and NO_x, respectively, as compared to a decrease of 3% and 7% in CbW.

3.1.2. Simulated PM_2.5 levels

We compare our BASE simulation results with observations, and find that our simulations capture the day-to-day PM_2.5 variation for four representative months in 2010 in Beijing (figure 2(b)) and other regions in East Asia (online supplementary figures 1–3). The BASE PM_2.5 concentrations are generally higher in eastern and southwest China, and during autumn and winter (figure 2(a)), which are largely affected by seasonally varying meteorological conditions.

Zoom In Zoom Out Reset image size

In the eastern provinces—the ones targeted for pollution reduction—electricity transmission irrespective of fuel source results in a 2–3 μg m⁻³ (or 2%–7%) reduction in the annual mean PM_2.5 concentrations, roughly 1 μg m⁻³ greater than the reduction achieved in CbR (figure 2(c)). Although switching from coal to hybrid power transmission does not further improve eastern air quality, adding zero-emission renewable capacity would avoid the PM_2.5 increase in some northern exporting regions observed in CbW. Since downwind regions of the eastern importing provinces can also benefit from the air quality improvement there, even in CbW we find the PM_2.5 levels in the exporting regions located in the northeast or surrounded by the importing regions are slightly lower than their BASE levels. We also find small PM_2.5 reductions in southern and central China in all three scenarios, because these regions are downwind of the eastern provinces for the most time throughout the year (see online supplementary figures 6 and 7 for monthly PM_2.5 results and wind fields). Such results are robust under meteorological conditions favorable for pollution formation and accumulation (see online supplementary figure 10). Since our BASE simulation does not show systematic biases, the general pattern of the PM_2.5 reduction in each scenario and the ranking of the three scenarios should be robust.

3.1.3. Air-pollution-related deaths

We estimate the national total annual premature deaths associated with the exposure to outdoor PM_2.5 to be 0.85 million in BASE (comparable to 0.86 million in GBD study [34]). Compared with BASE, Hybrid avoids 16% more premature mortalities nationally than CbW (∼16 000 and 14 000 cases, respectively). Most of the additional avoided deaths occur in the exporting regions, due to slightly lower PM_2.5 concentrations there when more renewable generation is utilized for transmission. In comparison, the two coal-based scenarios yield similar scale of avoided deaths (only 6% higher for CbW than CbR) (figure 3(b)). Therefore, at the national level, CbW has modest additional health-related air quality benefits compared with CbR despite the fact that Chinese planners have focused on wiring coal precisely for its air pollution advantages. Instead, CbW shifts the geographical pattern of mortality impacts by avoiding more deaths in polluted eastern cities but fewer deaths in the rest of the country compared to CbR. Such results are largely driven by the concavity of the relative risk functions. While a small PM_2.5 decrease in less polluted, southern regions results in nontrivial reduction in mortality risk and hence premature mortality, the flattening of RR functions is relevant at the BASE PM_2.5 levels in many eastern provinces, leading to smaller mortality risk reductions there (figure 3(a)). Applying linear RR functions would increase the magnitudes of avoided deaths in the energy transfer scenarios and the differences across them (online supplementary figure 11).

Zoom In Zoom Out Reset image size

Besides PM_2.5, ground-level ozone is also associated with negative health impacts, but with a smaller mortality risk compared with exposure to PM_2.5 [35, 36]. We find a small increase in eastern ozone concentrations in the three scenarios because lowered NO_x emissions increase ozone concentrations due to non-linear chemistry. However, the mortality impacts associated with such a small change in ozone exposure is unlikely to be substantial (More details in online supplementary figures 4 and 8).

3.2. Carbon emissions

With zero-emitting renewable energy displacing coal power generation, a net annual reduction of roughly 340 million tons of CO₂ emissions is obtained in Hybrid, equivalent to a 4% decrease in national total all-sector carbon emissions. By contrast, CbW and CbR have little effect on carbon emissions, as reductions only result from efficiency improvements in the new coal units relative to the displaced ones. The scale of the net reduction in CbR and CbW is nearly identical, as the same new coal power generation replaces old coal generation in both scenarios though in different locations (figure 4). Hybrid reduces carbon emissions three times more than either coal-based scenario, suggesting large global climate benefits of utilizing transmission lines to deliver a hybrid of renewable and coal power instead of coal power alone. The importance of the power mix in determining the carbon implications of electricity transmission has also been found in prior studies on carbon emissions embodied in inter-regional transmission in China [11]. Although our results do not consider transmission losses, the effect of transmission losses on carbon emissions are generally small as shown in our sensitivity analyses (online supplementary figure 9).

Zoom In Zoom Out Reset image size

Our evaluation of the air quality and climate benefits of the three energy transfer scenarios is subject to four main sources of uncertainty: transmission losses, meteorological variation, shape of the concentration-response relationships, and potentially a lower utilization rate of the transmission lines than we assumed in our Hybrid scenario. By conducting sensitivity analyses (section 5 in the online supplementary Materials), we find our main conclusion that Hybrid-by-wire results in greater air quality and climate benefits than coal-based options is still robust.

Here we discuss the cost-effectiveness of various energy-by-wire strategies by combining available cost data with the national total air quality and climate benefits evaluated in the previous section. We report the economic values in inflation-adjusted 2010US$.

a) CbW vs. CbR

When coal is the fuel choice, our analysis suggests similar climate benefits and only slightly larger health-related air quality benefits from CbW as compared to CbR at the national scale (6% greater reduction in national total premature deaths in CbW than CbR). On the cost side, CbR incurs coal transport costs, which lead to higher delivered coal prices for coal power generation in the eastern importing provinces than inland exporting provinces. As such, based on province-specific levelized cost of coal-fired electricity (online supplementary table 7), we find that the annualized coal power generation costs are US$4bn higher in CbR than CbW due to coal transport costs. The transmission cost to enable CbW, however, is difficult to estimate, due to limited existing cost data for transmission lines that use advanced ultra-high-voltage (UHV) technologies. Davidson et al estimated the UHV transmission cost to be 0.095 (low: 0.076, high: 0.126) RMB kWh⁻¹ and 0.111 (low: 0.088, high: 0.147) RMB kWh⁻¹ from Northeast or Northwest China to the Sanhua regions (the combination of Central, East and North China), respectively. Such estimates are also comparable with the current price for inter-regional transmission in China determined by the central government (roughly 0.12 RMB kWh⁻¹) [38]. Therefore, we estimate the annual total transmission costs to be roughly US $6.7bn (low: $5.3bn, high, $8.9bn; more details in online supplementary table 8 and 9). This indicates that at present it may cost more to transmit coal power than to transport coal by rail followed by electricity generation. Our evaluation hence provides no evidence that the CbW strategy would be more cost-effective than CbR in curbing national total air pollution impacts or global climate impacts, although in the targeted eastern provinces, CbW indeed reduces air pollution and health impacts more than CbR. Since UHV transmission is a new and advanced technology, future UHV costs are likely to decrease as learning progresses, which will increase the competitiveness of the CbW strategy. Our results are consistent with previous findings that cost-effectivness is determined by a combination of the investment costs for new infrastructure and the quantity of energy transported [16].

b) Hybrid vs. CbW

To evaluate the relative cost-effectiveness of transmitting a hybrid of renewable plus coal power as compared to coal power alone, we estimate that in 2010 the total power generation cost in Hybrid to be US$11bn higher than CbW, due to higher generation costs for wind than coal (average levelized cost of electricity for wind and coal in exporting regions in 2010: US$72 MWh⁻¹ and US$36 MWh⁻¹, see online supplementary table 7). Here we neglect the potential costs to manage the transmission of variable wind power supply as compared to reliable coal power supply. It is likely that these costs have decreased since 2010, however.

On the benefit side, we find that Hybrid reduces 16% more air-pollution-related deaths (∼2000 cases) and three times more carbon emissions (∼260 million ton) than CbW. However, monetizing these air quality and climate benefits is controversial and is often based on the value of a statistical life (VSL, a measure of people's willingness to pay to reduce their mortality risk) and social cost of carbon (SCC, the marginal global societal cost of emitting an additional ton of carbon emissions), respectively. The VSL estimates vary substantially across countries, income groups, age groups [39], and implicitly involve some value judgment [40]. The SCC estimates range from $20 to higher than $200 ton⁻¹ in the literature [41–45], and may still be underestimated due to the high social discount rate being applied [41] and the omission of other damages [42, 46]. Therefore, instead of choosing specific VSL or SCC values, we calculate the breakeven VSL to favor Hybrid over CbW, above which the monetized value of the additional air quality benefits in Hybrid can justify the more expensive costs for renewable generation. We also estimate the breakeven carbon price under a range of VSLs above which the monetized air quality and climate benefits would justify the higher renewable costs in Hybrid.

Considering only the air quality benefits, a VSL higher than US$5.0 m (confidence intervals due to RR functions: 3.5 m, 10.2 m) would favor Hybrid over CbW. Such breakeven VSL is more than twenty times higher than China's present-day VSL estimated from survey studies (low and high estmates $90 000 [47] and $250 000 [48]), or three times as high as the estimate obtained by adjusting the U.S. 2005 VSL with income elasticity (high estimate of $1.6 m for 2005 VSL in West et al 2013 [49]). Considering the climate benefits alone, we find a breakeven carbon price of $43 ton⁻¹ to favor Hybrid (figure 5). Considering combined air quality and climate benefits, the breakeven carbon price would only be slightly lowered under VSL estimates based on surveys ($40–42 ton⁻¹).

Zoom In Zoom Out Reset image size

To what degree do these benefits from using renewables accrue nationally? A full analysis should look not just at national air pollution benefits but also the benefits from lessening climate impacts that may occur domestically. That requires decomposing SCC into national and global impacts—a task not done much in the literature. However, Anthoff et al 2011 [45] found that 48% (or 31%) of total global climate damages will occur in China under a 0.1% (or 1%) discount rate. That suggests a global SCC of around $80 ton⁻¹ (or $120 ton⁻¹) would yield a Chinese portion of costs in the $40 ton⁻¹ level that we find favor Hybrid. This range of global SCC is consistent with the values found in the extensive western literature on the global impacts of carbon emissions [41–45].

Therefore, based entirely on plausible national benefits—mainly improvements in public health and to a lesser degree lower impacts from climate change—the strategy of transmitting a hybrid mix of renewables and coal already appears to be the favorable fuel choice. The Hybrid strategy will likely become more attractive in the future with a decline in renewable technology costs and a growing economic valuation of the benefits. The drop in wind technology costs in China from 2010 to 2014 [50, 51] could already lower the wind LCOE in exporting regions by 6%, hence narrowing the cost difference between Hybrid and CbW and lowering the breakeven carbon price (figure 5). People's willingness-to-pay to reduce mortality risk (i.e. the VSL) may also increase with growing income level in the future, which consequently favors Hybrid.

We highlight three policy implications. First, coordinating transmission planning with renewable energy deployment is critical to maximize combined air quality and climate benefits from energy by wire strategies. In China for example, wind and solar resources are abundant in northern and western regions, while air pollution is severe in eastern population centers. Using electricity transmission to connect renewable production areas with highly populated and polluted regions can better exploit renewable resources in remote areas, integrate variable with dispatchable resources [55], reduce carbon emissions, and maximize air quality and health benefits [53]. As many countries also need to expand transmission to support a scale-up of renewable energy [5, 17], we suggest that grid planners, acting in the national interest, consider the air quality implications of transmission capacity investment in order to increase both potential health co-benefits and carbon mitigation efforts.

Second, we highlight that work of this type—which aims to quantify the national benefits of policies that could have global consequences—requires future improvements in calculating national and global benefits in economic terms. The integrated assessment method in this analysis provides a rigorous evaluation of the health-related air quality impacts and carbon emissions. However, monetizing the impacts currently rely on VSL estimates, which in emerging economies are particularly hard to pin down, as well as concepts such as the SCC, which co-mingles national and global benefits. More robust economic evaluations are needed for reliable cost benefit analyses [54].

Third, while we focus on air quality and climate impacts in this analysis, long-distance transmission may pose other local environmental impacts especially in the electricity exporting regions. For instance, relocating coal power generaion to arid western regions may exacerbate water scarcity and extensive development of hydropower may have major impacts on local ecosystem. We thus suggest that grid planners consider the overall impact of long-distance electricity transmission on the environment at regional, national and global scales.

We thank Eric Larson, Guangjian Liu, Michael Oppenheimer, Mark O'Malley, Fabian Wagner, Robert Williams and Tong Zhu for valuable discussions and the GIS support of Tsering W Shawa.