Contribution of Offshore Wind to the Power Grid: U.S. Air Quality Implications

Abstract

Offshore wind is an established technology in Europe and Asia, but it has not yet gained market share in the United States. There is, however, increasing interest in offshore wind development in many coastal regions of the United States. As offshore wind grows in those regions it will displace existing and future electric generation assets, which will lead to changes in the emissions from the electric power sector. This research explores combinations of two electric sector drivers, offshore wind capital costs and carbon dioxide (CO₂) caps, to measure the changes in the energy mix and quantify offshore wind’s impact on electric sector emissions. An energy system modeling approach is applied, using a nested parametric sensitivity analysis, to generate and explore potential energy futures and analyze the air quality and greenhouse gas emissions benefits of offshore wind as an energy source. The analysis shows that offshore wind capacity was added due to cost reductions more than CO₂ cap stringency, though both increased capacity additions. Capacity varied more by CO₂ cap stringency at higher prices and less at lower prices. CO₂ mitigation led to reductions in all five emissions investigated, regardless of offshore wind cost. Offshore wind-specific reductions were only consistent across all CO₂ caps for CO₂ and methane (CH₄), though offshore wind-specific reductions were found for all emissions in the absence of CO₂ caps. Results are presented nationally, analyzing the differences in adoption of offshore wind and how this technology provides a broader range of emission reduction options for the power sector.

1. Introduction and objectives

1.1. Objectives

Offshore wind (OSW) is a renewable energy resource available over coastal and great lake waters. Its low variability and low uncertainty paired with its proximity to large population centers makes it a prime candidate for electricity production. In the U.S., approximately 40% of the population lives on the coast, and this coastal population has increased by 40% since 1970 [1, 2]. Offshore winds blow relatively consistently and often peak in tandem with daily demands. While electricity demand per capita has declined with energy efficiency improvements and behavioral change, total consumption continues to grow. As the power grid grows to meet demand and shifts as older generation resources retire, new technologies will come online to fill the gaps. Research that has been conducted to analyze how OSW will contribute to grid mix changes within the U.S will be explored in sections 2.1 and 2.2, but that research has focused either on resource assessment or technical solutions to grid integration, leaving a gap in an integrated assessment of OSW within the context of the power grid. Thus, an energy system modeling approach is used here to generate and explore these potential energy futures, with the objective of assessing whether OSW will be part of the change, and if so, quantifying how it contributes to changes in electric sector emissions. Two drivers for OSW development are explored: (1) technology costs and (2) carbon cap stringency.

1.2. Technology Costs

The U.S. has an estimated 10,800 GW of OSW resource potential, 2,058 GW of which are technically feasible for development [3]. Though the resources are vast, it is a relatively expensive technology and only 30 MW of this potential has been realized. All five OSW turbines in the U.S. came online in late 2016 at the Block Island Wind Farm located off the coast of Rhode Island. Many factors contribute to the high cost, the most impactful being complex installation that requires highly-skilled instrumentation and labor at sea [4, 5]. Supply chains for these resources are not yet developed in the U.S. Distance to shore and depth of water add further to these costs. Lastly, the electricity produced must make its way to shore through sea-worthy and costly transmission lines [6]. This results in a high cost for OSW as compared to other technologies. However, there is a great deal of potential for declining costs for OSW. As capacity expansion in the power sector is highly sensitive to cost, this driver captures one of the main barriers to OSW deployment.

1.3. Carbon cap stringency

Electricity generation produces several emissions, including but not limited to sulfur dioxide (SO₂), nitrogen oxides (NO_X), fine particulate matter (PM_2.5), methane (CH₄), and carbon dioxide (CO₂). These emissions vary in their environmental and health impacts, as well as their cost of mitigation. Currently, the electric sector contributes approximately 69% of SO₂ emissions, 33% of CO₂ emissions, 14% of NOx emissions, and 3% of PM 2.5 emissions. Methane emissions in the U.S. occur primarily through natural gas and oil production, transmission, storage, distribution, and processing, 36% of which is used in the electric sector [7, 8]. Federal programs already exist for the mitigation of SO₂, NOx, and PM_2.5 from this sector, such as the Acid Rain Program and Cross-State Air Pollution Rule, but not yet for CO₂. Twenty-four coastal and Great Lakes states and Washington D.C. have instituted Renewable Portfolio Standards (RPS) and twenty have set greenhouse gas (GHG) emissions targets [9, 10]. Both types of policies incentivize the buildout of renewable and emissions-free generation resources, for which OSW qualifies. These policies have already begun to change the energy landscape. Policies, paired with declining costs for terrestrial wind and solar, led renewables to account for 17% of electricity generation in the U.S. in 2018, an increase of 5% from 2017 [11, 12]. The 2019 Annual Energy Outlook (AEO) produced by the U.S. Energy Information Administration (EIA) projects that electric sector emissions will remain flat from 2020 through 2050, assuming there are no significant changes to laws and regulation [13]. Carbon mitigation stringency was explored to account for air-quality and environmental health regulations that would favor non-emitting sources of power generation beyond what already exists. This driver also helps to encapsulate the upward trend in states with GHG emissions targets and the stringency of those targets.

2. Background

2.1. OSW Resources and Implementation

OSW has been successful in Europe, with over 18.9 GW of installed capacity, though it has not yet taken off in the U.S. and globally. It has reached cost competitiveness with many other technologies, but terrestrial wind and solar photovoltaic (PV) still remain less expensive than OSW [14]. The U.S. has substantial OSW resource availability and significant research has been conducted to investigate the feasibility of harnessing these resources, specifically accounting for technological limits, land availability, risk factors, and economics [3, 15–17]. A recent approach developed by Dupont et al. works to include Energy Return on Investment (EROI), incorporating more robust econometric measures into the resource assessment [18]. The U.S. Department of Energy has done a preliminary assessment of interconnection for OSW and concluded that the U.S. (1) has sufficient OSW resources to build out a considerable amount of generation and (2) appropriate technologies exist for interconnecting large amounts of OSW energy to the U.S. grid, however realizing this potential will be both complex and difficult [19]. Adding a high penetration of variable renewable resources to the grid is a major complexity due to their non-dispatchability. Huang et al. developed control strategies for grid operators for highly OSW penetrated systems [20] that can be paired with research from Moller et al to harness spatial modeling for later-stage planning [21]. Also, there remains the issue of transmission of the OSW-generated electricity to shore, where the U.S. resources vary greatly by region in sea floor depth and distance to shore. This makes economic and location-based analysis difficult given the highly variable cost and requirements for transmission. Models for configuration and cost optimization of OSW transmission systems have been developed by multiple research teams, showing promising solutions for the vast technical difficulties [22–24]. The U.S. Bureau of Ocean Energy Management (BOEM), the agency responsible for OSW leasing areas, announced in June of 2019 that it would publish a “request for competitive interest” to build a transmission line for OSW off the coasts of New York and New Jersey [25]. A planned transmission system for OSW would promote long-term success and interest in the OSW market, and likely increase the chances of the U.S. market accelerating OSW development.

OSW is still at a relatively early stage of technology development, even in Europe, and prices are expected to fall further as the technology advances [26]. In 2017, the National Renewable Energy Laboratory (NREL) published a study assessing the economic potential of OSW in the U.S. that projected a decline of approximately 50% in the levelized cost of electricity (LCOE) of OSW, both in shallow and deep waters, by 2030 [15]. Supply chain and infrastructure advancements are key to these price declines, but the largest contributor is growth in the capacity of turbines. In 2016 the Block Island Wind Farm installed five 6 MW turbines, standing at a height of 590 ft. In 2019, the newest and largest turbine design, the GE Halide X, has doubled to a 12 MW capacity and is 863 ft tall, increasing its capacity factor by five to seven percent above the current industry standard [27]. Steep LCOE decline has precedent in the U.S. Since 2009, the LCOE of solar PV and terrestrial wind have declined 88% and 69%, respectively [28]. These declines reflect both technological advancement and economies of scale for these technologies. OSW expects to benefit from these factors, enhanced by industry and state government interest in investing in the technology. Daniel et al. projects that research and development will help reduce initial capital investment for OSW due to industry advancements in these areas [19]. However, OSW still stands as one of the most expensive generation resources available in the U.S., as seen in Table 1. While LCOE is a generalized measure used to compare technologies, it does not account for regional differences in fuel or natural resource availability, load curves, or dispatchability. Additionally, uncertainty and unpredictability in the assessment of wind resources can change the LCOE, as quantified by Mora et al. [29].

Table 1.

Estimated LCOE capacity-weighted average for new generation resources entering service in 2023 (2018 $/MWh)

Plant Type	Capacity Factor (%)	Levelized capital cost	Levelized fixed operation & maintenance	Levelized variable operation & maintenance	Levelized transmission cost	Total system LCOE	Total LCOE including tax credit
Dispatchable technologies
Conventional Combined Cycle	87	8.1	1.5	32.3	0.9	42.8	42.8
Advanced Combined Cycle	87	7.1	1.4	30.7	1	40.2	40.2
Advanced Combustion Turbine	30	17.2	2.7	54.6	3	77.5	77.5
Geothermal	90	24.6	13.3	0	1.4	39.4	36.9
Biomass	83	37.3	15.7	37.5	1.5	92.1	92.1
Non-dispatchable technologies
Wind, onshore	44	27.8	12.6	0	2.4	42.8	36.6
Wind, offshore	45	95.5	20.4	0	2.1	118	106.5
Solar Photovoltaic	29	37.1	8.8	0	2.9	48.8	37.6
Hydroelectric	75	29.9	6.2	1.4	1.6	39.1	39.1

Source: U.S.EIA AEO 2019 [13]

Despite the high price of OSW in the U.S., many states are taking steps towards incentivizing and implementing the technology. Recent state advancements in supporting OSW projects and supply chain include work in California [30], Connecticut [31], Maine [32], Maryland [33], Massachusetts [34], New Jersey [35], New York [36], Rhode Island [32], and Virginia [37]. The U.S. market is at a tipping point, with a project pipeline of 25,600 MW of OSW energy as of July 2019. Specific projects account for 4,864 MW of that capacity and the remaining 20,736 MW are comprised of undeveloped lease area. Of the project-based capacity, 4,831 MW is expected to be built and online by 2024 [32].

2.2. Environmental and Health Implications

Renewable energy has grown significantly in the U.S. as federal incentives have grown, states have adopted RPS and emissions reductions goals, and especially as costs have declined. A multitude of factors elicit greater amounts of specific technologies in different regions depending on natural resource availability, demand growth, age of existing generation assets, incentive policies, and many more factors. Little research exists for the impact of OSW to the grid mix in the U.S., but many other countries are taking initiative to assess OSW within their own contexts. These efforts have been focused primarily in European countries and China, as they were the first and most prolific adopters of OSW. Researchers have begun assessing the opportunities in China for OSW and other carbon-free technologies to offset carbon-intensive generation in the residential, industrial, and electric sectors [38–40] as well as more wholistic investigation into OSW planning for emissions reductions goals [41, 42], though these studies have not used an integrated assessment modeling approach. Similarly, international research has sought to quantify the ability of renewable technologies to contribute to emissions goals [43–45], and wind specifically [46]. This research has shown a direct correlation between adoption of renewable technologies and emissions reductions, accounting for demand and population growth. Electric sector emissions reductions contribute to benefits for the environment, human health, and climate, often referred to as “co-benefits”. Analyses have been done to quantify the positive human health co-benefits from sectoral emissions reductions [47, 48] and more general climate mitigation efforts [49, 50]. Additional research has quantified the environmental and climate co-benefits of low carbon technologies [51–53].

U.S.-specific research has estimated the environmental co-benefits of low-carbon pathways [54, 55], and has begun to investigate terrestrial wind specifically [56, 57]. In 2017, BOEM issued a report evaluating electricity system, environmental, and socioeconomic benefits offered by OSW [58], modeling a novel approach to quantifying more than costs and impacts when assessing new technologies. The U.S., while adopting very little OSW, has installed nearly 98 GW of terrestrial wind power [59], allowing for real-time analysis of emissions mitigation potential and the resulting impacts on and benefits for environmental, human health, and climate [60, 61]. As interest in OSW has grown, so has investigation into the co-benefits that the technology can bring at the local level. Two studies on the state of Michigan showed that both terrestrial and offshore wind could bring environmental, air quality, and quality of life benefits to the counties that would host or border the turbines [62, 63]. They found that emissions reduction benefits varied across pollutant and locality, but that overall reductions were likely across all pollutants. A similar study focused on Mid-Atlantic states found direct human health and climate benefits from OSW specifically, citing the greatest differences in benefits coming from locality and facility generation capacity [64]. It must be noted that OSW does not produce emissions from generation, however materials production, construction, and operation of OSW have their own emissions footprints, which are quantified in life-cycle analyses [65–67], and there are emissions consequences when variable renewable resources must be paired with dispatchable, fossil-fuel generators to meet demand when renewable generation drops [68]. These contextual emissions have not yet been worked into a larger systems-approach to emissions mitigation, nor are they accounted for in this study.

These studies demonstrate the direct effects and benefits of carbon-free electric generation technologies and pathways to their surroundings, but do not quantify the national impact that OSW would have on emissions. The power grid in the U.S., as in many other nations, is robust and does not necessarily emit pollutants at the point of end use. Often these systems are integrated, and electricity travels a long way. While the few studies at the local level for OSW benefit assessment in the U.S. are thorough, they do not assess the probable far-reaching impact of OSW, nor do they investigate how OSW might actually change the generation composition of the grid. U.S. governmental agencies are confident in the state of technological feasibility and development planning for OSW, validated by the ongoing success in Europe and Asia, however there still has been no assessment of what that development might look like, what factors will influence it, and how it will affect emissions nation-wide. The research presented here seeks to provide that assessment and characterize how OSW will impact the grid mix and influence electric sector emissions.

3. Materials and Methods

3.1. Methodology

This research uses an energy system modeling approach to explore potential energy futures that include OSW and analyze the resulting changes to the electric sector technology mix and associated emissions. The methodology developed in Loughlin et al. (2012) for assessing the breakthrough potential of emerging technologies for emissions mitigation was applied to OSW [69]. The Loughlin et al. methodology was designed to evaluate the potential of energy technology developments to yield a breakthrough in achieving GHG mitigation goals. It was applied more broadly in this research to evaluate the changes that OSW would elicit in the energy mix as CO₂ cap stringency and OSW costs change and the resulting grid emissions. The approach involves a nested parametric sensitivity analysis using the The Integrated MARKAL-EFOM System (TIMES) energy system model and a nine-region database representation of the U.S. energy system called the EPAUS9rT. Combinations of carbon cap stringencies and OSW cost curves create vastly different energy futures with comparably different emissions profiles. The study allows us to analyze, given the uncertainty of any one “future” scenario, the factors associated with the implementation of OSW and the benefits to system emissions associated with an increase in OSW power.

3.2. Model and Database

The TIMES model and EPAUS9rT energy system database, together, provide a comprehensive look at the U.S. energy system. The EPA’s Office of Research and Development has worked to develop the TIMES-EPAUS9rT modeling system to investigate wholistic energy system futures that optimize for lowest cost over all economic sectors [70]. This allows the use of a singular model to represent all energy system sectors instead of a piecemeal approach, showing the interplay and tradeoffs between sectors as scenarios change. The TIMES model is developed by the Energy Technology Systems Analysis Program (ETSAP), one of the longest running programs at the International Energy Agency (IEA) [71]. The TIMES model is a long-term energy system optimization model (ESOM) and is used for investment and operation decision support. It uses a bottom up approach with multi-year temporal resolution and user defined time slices to model for long time-horizons [72]. The time slices represented in the model are seasonal, peak, night, day am, and day pm. The TIMES model encompasses all steps from primary energy through processes that transform, transmit, distribute, and convert then to their final end-use. It includes a wide range of commodity and process-related variables such as total production, total consumption, and process flows. TIMES allocates costs across investment timelines for each of these steps.

The TIMES model is a linear program formulated using the GAMS modeling language that maximizes system surplus and minimizes system costs. This formulation, Equation 1, is called the total system cost [71].

$$NPV=\sum _{r=1}^R\sum _{y\in YEARS}{(1+d_{r,y})}^{REFYR-y}\times ANNCOST\left(r,y\right)$$ (1)

Where

• NPV is the net present value of total cost for all regions;

• ANNCOST(r,y) is the total annual cost in region r and year y;

• d_r,y is the overall discount rate;

• REFYR is discounting reference year;

• YEARS is the number of years for which the results will be modeled; and

• R is the group of regions in the study.

The TIMES objective is to minimize the total cost of the system, augmented by the cost of lost demand. All cost elements are discounted to a user-selected year [71]. The model produces the least cost energy system that meets all energy demands over the time horizon, adhering to the economic and engineering relationships between energies and their uses, resulting in the optimal mix of technologies and fuels at each defined period, as well as their associated emissions. TIMES has perfect foresight, and therefore is limited in making “real-world” decisions based on market conditions. All additional constraints can be added to the base TIMES objective function through the use of scenarios. The current EPAUS9rT database and its TIMES implementation uses 2010 as the base year and is calibrated to the present year, including implementation of scenarios that represent constraints in the U.S. energy system, such as federal emissions regulations and existing energy demands.

The EPAUS9rT database represents the U.S. by census regions, as can be seen in Figure 4, to capture big-picture regional characterizations. OSW’s representation in this model is at this region-level and accounts for the geographic and economic variability of the resource. The OSW resources within the model span every census region except for Region 8 (Mountain West), due to no offshore resources being directly accessible from this territory. For each of the remaining eight (8) regions, the model represents technologies for OSW spanning the following characteristics: water depth (shallow and deep), wind class, and cost class. Capacity factors vary by time of day, season, technology, and region. For each OSW installation, a lifetime of 30 years was assumed. Though some assessments apply a 25-year lifetime, we have tried to account for technological advancement and lifetime extensions, as was shown with the earliest offshore wind farm, Vindeby. Vindeby operated for 26 years in Denmark, and since its commissioning in 1991 there has been substantial progress made in operating capacity and durability [73]. Thus, we have decided to use 30 years as the lifetime for offshore wind farms. The model sets a capacity bound for type of OSW in each region based on technical feasibility. Not considered in the model is the practical timeframe in which all the available OSW development area will become available as BOEM, must assess outer continental shelf areas for leasing potential before they may be developed. Additionally, curtailment of renewables is not considered because the model builds only the capacity needed to meet demand based on the resource availability across regions and time slices.

Figure 4.

Cumulative OSW capacity per region in the 70% Cost Reduction & 70% CO2 Emissions Reduction Scenario

3.3. Scenarios

Figures 1a and 1b are graphical representations of the scenarios for CO₂ caps and OSW cost curves implemented in this study.

Figure 1.

a. CO₂ Cap and b. OSW Cost Curve Scenario Construction

CO₂ cap scenarios were constructed to linearly reduce emissions from 2010 to 2050 by a percentage of 2010 electric sector CO₂ emissions and were implemented as an electric sector CO₂ upper bound (CO₂ cap) (Figure 1a). The 2010 CO₂ emissions that serve as the baseline for reduction scenarios were calculated endogenously in a reference case run, using the TIMES-EPAUS9rT modeling system. The reference CO₂ cap scenario assumed no carbon mitigation requirement and reaches approximately a 25% CO₂ reduction by 2050 due to existing state-level policies implemented in the reference case, such as renewable portfolio standards, and natural turnover of aging generators. Not included in the reference case are state-level emissions reduction targets. Six additional scenarios were constructed, starting with a 30% reduction and subsequently increasing the required carbon mitigation percentage by 10%, until an 80% carbon reduction is achieved by 2050. Based on existing state CO₂ reduction targets, nation-wide emissions are projected to continue to decrease, with some individual states setting aggressive targets, such as an 80% reduction by 2050 in New York and Maine and carbon neutrality by 2045 in California. The U.S. EIA’s AEO projection of electric sector CO₂ emissions falls at around 40% reductions from 2010 to 2050, accounting for many of these state policies [13]. The speed of emissions reduction and success of these targets relies on the success of these policies, and that is still uncertain. Additionally, more states could pass emissions reduction policies in the coming decades. These uncertainties are captured by the 7 emissions scenarios modeled.

Cost curves for OSW were constructed to linearly decline from 2015 to 2035 by a percentage of current costs and hold steady until 2050 (Figure 1b). Capital expenditure (CAPEX) was calculated using base overnight costs presented in the AEO [13]. The baseline cost reduction scenario assumes a 20% cost decrease, representing the minimal total technological learning by 2035 for OSW found in the 2019 AEO[74]. The EPAUS9rT database assumes technological learning for solar and terrestrial wind as well, though their values are quite modest as a lot of the technological advancement and economies of scale have already been achieved. Five additional cost curves for OSW were constructed at 10% intervals, spanning from a 30% to an 80% reduction in the cost of OSW by 2035. The 50% cost reduction curve most closely represents NREL OSW LCOE projections, accounting for growth in the capacity of turbines and supply chain and infrastructure advancements [15].

The reference case for this study (Ref) is represented by both the reference CO₂ cap and OSW cost curve scenarios, applying no CO₂ cap to the electric sector and assuming only the minimum technological learning and cost reductions for OSW.

4. Results and Discussion

4.1. Buildout and Capacity

A total of 49 scenarios were created, representing the combination of all cost curves and CO₂ caps. The model results show that OSW was only built out in 24 scenarios, as represented in Figure 2. Until the cost of OSW is reduced by 40%, it is not economically viable, and at 40% cost reductions it is only built in the most stringent carbon cap scenario. At a 50% cost reduction 21 GW of OSW is built by 2050 in the 80% carbon cap scenario, though only 11 GW is built in the 70% carbon cap scenario. At a 60% cost reduction and above, OSW is built in all carbon cap scenarios, with the largest capacity coming at the highest cost reduction across all carbon cap scenarios.

Figure 2.

Total OSW Capacity in 2050

OSW’s sensitivity to carbon cap stringency is well defined across all technology costs. More stringent carbon caps elicit the buildout of OSW sooner, and at larger capacities (Figure 3). Additionally, the less expensive OSW becomes, the more capacity is built out each time period, regardless of carbon cap stringency. Overall buildout of OSW varied across regions, with no capacity added in Regions 4 and 6 due to minimal resource availability and comparably higher costs and Region 8 due to no resource availability (Figure 4). The EPAUS9rT database does not have OSW availability for Region 8 because there is no coastline, and Regions 4 and 6 have very little resource availability. Of the regions where OSW was built, Regions 9 and 3 elicited the largest buildout and Region 7 the smallest, though the differences also varied between scenarios. Regions 1, 2, and 3 saw consistent buildout throughout all scenarios.

Figure 3.

Total OSW Capacity by Scenario

OSW, as a new technology, can be built to replace existing generation or to add capacity as electricity demand grows. All scenarios exhibit electricity demand growth over time, but the degree of the growth varies between scenarios. CO₂ caps limit the increase in total electricity generated as they become tighter, showing that the carbon constraint affects demand and electricity end uses (Figure 5). As OSW gets less expensive, however, total electricity production grows, compensating for and even increasing the total output over the reference case. When OSW is the least expensive and there is no carbon cap, total electricity production in 2050 is 9% greater than when it is most expensive with a stringent carbon cap. Across the tightest carbon cap scenarios, OSW is still able to elicit a 5% increase in total electricity production when it is least expensive.

Figure 5.

Total electric sector electricity production in 2050

As the carbon cap gets tighter, there is a tradeoff between the electric and industrial sectors. The industrial sector both consumes electricity provided by the grid and produces its own using combined heat and power (CHP). The CO₂ cap scenarios only apply to the electric sector and do not include emissions from industrial CHP. Therefore, as the CO₂ cap tightens, there is a small shift from industrial grid electricity use to industrial CHP electricity production, also shifting those emissions and increasing overall industrial sector emissions (Figure 6). As OSW costs decrease, that shift lessens with industrial grid electricity consumption increasing and CHP activity decreasing. This suggests that adding cost-effective OSW can minimize the need for the industrial sector to utilize CHP in scenarios with tight electric sector carbon caps. The increase in CHP electricity production and emissions are minimal in comparison to the large reductions seen in the electric sector as a whole as CO₂ caps tighten.

Figure 6.

Industrial sector CHP and Grid electricity use

4.2. Grid Mix

The reference case, shown in Figure 7, shows the grid mix changes through 2050 absent OSW cost reductions and an electric sector CO₂ cap. This figure also shows grid mix changes in two additional scenarios, representing a mid and high OSW capacity. Figure 8 shows the differences in electricity production for each technology between the indicated scenarios and the reference case (Figure 7), with net increases above the dotted red line and net decreases below. In most scenarios where OSW is built, its deployment displaces coal, natural gas, terrestrial wind, and solar PV, as seen in Figure 8, though the technologies displaced vary between scenarios. When OSW cost reductions are only 50%, little OSW is built. In order to meet the increasingly stringent CO₂ caps, solar, terrestrial wind, and coal with carbon capture and storage (Coal CCS) are built and displace the existing coal and new natural gas built in the reference case. At a 60% cost reduction, more OSW is built, displacing what would have otherwise been new solar, terrestrial wind, or coal CCS. As costs decrease to 70% and 80%, almost all added capacity is OSW, as it becomes less expensive than other carbon-free electric generation resources.

Figure 7.

Reference case electricity production by technology

Figure 8.

Capacity additions and retirements in relation to the reference case

In scenarios where OSW costs are low and new capacity is high, natural gas is the most displaced technology, whereas in higher OSW cost and low capacity scenarios more coal is retired. Natural gas makes up a large market share of the 2050 power grid in all scenarios regardless of OSW buildout (Table 2), but natural gas capacity additions are dramatically reduced as the cost of OSW falls (Figure 8). Coal sees a similar displacement when OSW is built out in low quantities, though as OSW costs decline and capacity increases coal retirements slow and more existing coal remains over time. The high costs of carbon-free generation additions to meet the cap increases the total system cost. In order to offset these costs, coal remains online longer instead of being replaced with more quickly dispatchable and lower-carbon natural gas, though the overall system emissions decrease to meet the cap. This shows the tradeoff between building new carbon-free but non-dispatchable capacity and needing to meet demand at all times.

Table 2.

Percent market share by technology in 2050 (Technology TWh/Total electric sector TWh)

		Cost Reduction (%)						Cost Reduction (%)
Technology	CO2 Cap (%)	50	60	70	80	Technology	CO2 Cap (%)	50	60	70	80
Offshore Wind	BAU	0	2.3	7.1	24.2	Coal	BAU	17.9	17.7	17.2	15.6
	40	0	3.1	9	24.2		40	11.9	12.1	13.1	15.6
	60	0	4.8	17.1	31.5		60	5.1	6.2	8	9.2
	80	2.1	9.7	25.5	38.2		80	0	0	0.9	3.6
Terrestrial Wind	BAU	8.3	6.7	5.8	4.5	Coal CCS	BAU	0	0	0	0
	40	9.9	8.1	6.3	4.5		40	0	0	0	0
	60	16.8	14.6	9.6	5.8		60	1.3	0.6	0	0
	80	22.3	19.1	12	7.9		80	3.8	3.4	2.5	1.6
Solar	BAU	9	8.4	7.7	4.7	Natural Gas	BAU	41.5	41.7	39.1	27.7
	40	12.9	12	10.3	4.7		40	41.6	41	37.8	27.7
	60	18.3	17.6	14.2	7.5		60	34.1	31.6	26.9	22.5
	80	26.5	23	16.9	13.5		80	20	19.8	17.6	11.2
Hydro	BAU	5.5	5.5	5.5	5.8	Nuclear	BAU	17.1	17	17	16.8
	40	5.5	5.5	5.6	5.8		40	17.5	17.4	17.2	16.8
	60	5.7	5.8	5.8	5.8		60	18.1	18	17.7	17.1
	80	6.3	6.2	5.9	5.8		80	18.2	18.1	17.9	17.4

Over all scenarios, the largest market share that OSW achieves is 38% in the lowest cost and highest carbon cap stringency scenario. In all lowest cost reduction scenarios, OSW gains significant market share, but as costs increase that market share is more sensitive to the stringency of the CO₂ cap.

Additionally, as OSW becomes less expensive than solar and terrestrial wind at around 60% OSW cost reduction, these renewable technologies retain marginally less market share than they would have in the reference case. However, the total contribution of renewables increases as OSW costs decrease across all CO₂ cap scenarios because OSW capacity additions equal or surpass those of other renewable sources, as seen in Figure 9.

Figure 9.

Percent of electric sector production from renewable technologies. Renewable technologies include solar, terrestrial wind, and OSW.

4.3. Emissions Impacts and Sensitivities

Electric sector CO₂ emissions constraints similarly constrain SO₂, NO_x, CH₄, and PM_2.5 because they are cogenerated during fossil fuel combustion. The addition of OSW to the grid mix and the changes that it elicits vary the degree to which these other emissions are reduced. As shown in Figure 10, all pollutants saw a significant reduction in emissions, with the greatest overall reduction in SO₂ and PM_2.5. There is a clear trend showing greater reductions in all emissions as carbon mitigation stringency increases. For CO₂, SO₂, NO_x, and CH₄, tighter carbon constraints lead to emissions reductions surpassing the reference case beginning no later than 2030. This does not hold for PM_2.5, however, as OSW costs cause more variation in the pace at which this emission is reduced with fewer natural gas additions and slower retirement of existing coal. Due to the tradeoff between OSW and the slowing of coal retirements, PM_2.5 emission reductions do not outpace the reference case when OSW is less expensive and gains larger market shares.

Figure 10.

Electric sector emissions reductions

OSW-specific pollutant mitigation potential is less consistent across pollutants. There is, however, a clear trend through all of the reference CO₂ emissions scenarios. In all reference cases with greater OSW cost reductions, the greater the drop in cost of OSW the greater the reduction across all emissions. The OSW-specific reduction potential is strongest for CO₂ and CH₄, assuming no applied CO₂ caps, as well as across increasing CO₂ cap stringency. For CH₄ mitigation, OSW cost decreases are able to elicit larger reductions across all CO₂ cap scenarios, showing strong promise for OSW-specific CH₄ mitigation potential. The greatest potential is seen absent a CO₂ cap or with the most stringent 80% CO₂ cap, with lesser reduction with more moderate CO₂ caps. For CO₂ mitigation, OSW cost decreases are able to elicit larger reductions across all CO₂ cap scenarios, also showing strong promise for OSW-specific CO₂ mitigation potential. The greatest potential is seen absent a CO₂ cap, with diminishing reduction potential as CO₂ caps tighten.

OSW-specific reductions are achieved for PM_2.5, SO₂, and NO_X absent a CO₂ cap, but are not achieved across all other cap stringencies. There is a clear tradeoff between the reduction of these emissions and decreasing OSW costs. Due to the delayed retirement of coal-powered generators and replacement by more efficient and less polluting natural gas-powered generators when OSW comes online in larger capacities, the reduction of these emissions can be halted. With moderate CO₂ caps, PM_2.5 and NO_X only see OSW-specific reductions absent a CO₂ cap and with the most stringent CO₂ cap. The spread between emissions trajectories as OSW costs decrease is much greater for PM_2.5 than NO_X, showing that PM_2.5 emissions are more sensitive to OSW deployment. SO₂ emissions greater than in the reference case as OSW costs decrease whenever there is a CO₂ cap implemented, regardless of its stringency. This is due to coal-powered generators high SO₂ emissions and the tradeoff between OSW and coal retirements. Thus, OSW-specific reduction potential for SO₂ is unlikely to be achieved, regardless of scenario.

These results show that OSW-specific reduction potential lies mostly with GHG emissions and less so with the air quality indicators PM_2.5, SO₂, and NO_X. If substantial CO₂ reduction goals are met, however, OSW has a greater role to play in the reduction of these emissions, even when total electricity production increases.

Conclusions

The pipeline for offshore wind (OSW) development in the United States (U.S.) is growing and the research about the technical feasibility is robust, but still missing is the forward-looking research to assess how this technology will fit in, what will be displaced and when, and how it might affect U.S. air quality and emissions goals. This research characterizes OSW within this context to show how the grid’s generation mix will change as OSW costs decrease and electric sector carbon emissions are reduced. Costs for OSW will need to decrease by at least 40% before OSW is built, and 60% to compete with other technologies, both renewable and fossil fuel, for a sizable market share. At each price point, more OSW capacity is built as CO₂ caps tighten. Both study parameters have a positive effect on capacity buildout as they increase, with costs having the greatest positive impact. As carbon cap stringency increases, natural gas and coal are displaced at higher rates. As OSW costs decrease, it is better able to compete with other renewable technologies, as well as new natural gas and existing coal, leading to a greater displacement of new natural gas and slower growth of solar and terrestrial wind. Despite slower solar and terrestrial wind growth, OSW capacity additions lead to an overall increase in renewable contributions to the grid.

CO₂ caps elicited consistent reductions in CO₂, SO₂, NO_x, PM_2,5, and CH₄ emissions and OSW capacity in CO₂ and CH₄ emissions, though to a smaller degree than those from the CO₂ caps. SO₂ sees significant reductions in all scenarios, but in the tighter carbon cap scenarios, the range of SO₂ emissions reductions varies. The scenarios with lower OSW costs and higher penetration of OSW elicit less SO₂ reductions than the scenarios with higher OSW costs. Higher capacities of OSW, especially in the 80% cost reduction scenarios, require a greater capacity of dispatchable resources, slowing coal retirements by keeping them running longer than they would otherwise. Thus, this tradeoff impacts the potential for OSW to lower SO₂, PM_2,5, and NO_x emissions. OSW capacity elicits more pronounced and consistent reductions in all emissions at the least stringent CO₂ caps as there is no forced tradeoff between OSW and coal. Some of the emissions reductions in the electric sector are offset by gains due to end-use sector electricity production. Under tight CO₂ caps, the industrial sector consumes less electricity from the grid, where the CO₂ cap is applied, and increases its own electricity production with CHP, increasing industrial emissions but still significantly reducing emissions overall. The benefits of this systems approach can be seen in the identification of these tradeoffs. OSW is a new technology to the U.S. and its potential energy system effects cannot be fully analyzed unless the system can be seen as a whole. With emerging technologies, such as electric vehicles and hydrogen fuel, electricity is less isolated than in the past, operating more as a traditional fuel as the economy is electrificed and generation resources diversify. This research characterizes the role that OSW plays in this system, as well as the potential it holds to reduce emissions.

Further research might apply this methodology to a database that includes the OSW mandate policies of U.S. states and incorporates technology learning curves based on the current pipeline, assuming it comes to fruition. The technology landscape and cost reduction pathways are not yet realized, and state policies are likely to shape these as OSW begins to be built. Changes to emerging technology costs, and thus electricity costs, are likely to affect end-use electrification and cross-sector tradeoffs. Focusing research in this area could assess system-wide emissions outputs and sensitivities to the model parameters. The emissions reduction potential of OSW was analyzed at a national level in this study, but the TIMES-EPAUS9rT modeling system can provide results at the census region level. CO₂ caps could instead be applied at the regional level to better mimic state RPS and emissions goals. Extending the study to quantify emissions reduction benefits for environmental and human health would provide additional assessment of this emerging technology, at a national or regional level.

Notes on Modeling

The TIMES model [71] was implemented using the VEDA FrontEnd and BackEnd software suite and the EPAUS9rT database. OSW representation was developed for the database using the National Renewable Energy Laboratory’s Regional Energy Deployment System (ReEDS) model and database. No additional transmission expansion or offshore transmission system was modeled. This version of the EPAUS9rT database can be made available upon request by contacting author Carol Lenox (lenox.carol@epa.gov).

Highlights.

• Offshore wind capacity is modeled for the U.S. based on capital cost and CO₂ cap scenarios

• Natural gas and coal generation are most displaced by offshore wind

• Offshore wind additions can reduce electric sector CH₄ and CO₂ emissions

• Cost is the chief barrier to achieving these emissions reductions