In November 2017—just in time for the UN climate meeting in Bonn—the EU reached an agreement to reform and further tighten its Emissions Trading Scheme (ETS). In September 2017, French President Emmanuel Macron argued for higher carbon prices of at least EUR25-30/t
, while the Dutch government communicated a plan to impose a carbon price floor of EUR18/t starting in 2020
. The oil and gas industry also seems to agree; in 2016, the CEOs of some of the biggest oil and gas companies argued for the introduction of carbon prices to decarbonize the power sector by encouraging coal-to-gas switching
However, the actual impact of carbon prices on gas consumption is not straightforward. Carbon prices favor gas versus coal and can make gas the fuel of choice for baseload power. At the same time, carbon prices penalize gas versus increasingly competitive renewables. What is the “optimal” carbon price to increase gas consumption and to what degree would that switch reduce emissions? How does the effect of carbon prices on gas vary across markets and over time as the economics of renewables and storage evolve?
McKinsey Energy Insights develops an annual Global Energy Perspective (GEP). Using the GEP Power Model, we assessed how carbon prices can shift the power generation mix. We analyzed scenarios of multiple carbon price levels in different markets, modelling the effects on the generation mix as well as on emissions and power prices.
We found four key trends in all markets:
- There is a carbon price that maximizes gas consumption by being sufficiently high to displace coal and sufficiently low to avoid the displacement of gas by renewables in combination with storage.
- This “optimal price” from a gas perspective depends heavily on country-specific conditions. For gas consumption, country-specific carbon prices could be advantageous over a global price.
- The optimal price decreases significantly over time as the economics of renewables continue to improve. Gas tends to benefit from higher carbon prices (above $30/t) until 2025/30, accelerating the shift from coal to gas. Beyond 2030, gas is penalized by (nearly) any level of carbon prices.
- With regards to emissions, the shift from coal to gas is a significant abatement source in most countries. High carbon prices in the short run (until 2020) can shift generation and achieve substantial emission reductions by 2030 (e.g., prices of at least $30/t after 2020 would reduce emissions in the power sector by over 40 percent in 2030 for Germany).
Local conditions shape effects of carbon prices on natural gas
Higher carbon prices make coal less attractive, yet the impact on natural gas strongly depends on the local conditions. Breaking down the four above-mentioned trends, we uncovered six factors that affected the degree to which natural gas is favored when carbon pricing is introduced
- The carbon price changes the competitiveness of natural gas in comparison to other sources; as the carbon price rises, gas gains competitiveness over coal, but it also loses competitiveness over RES
- The existing power generation and capacity mix in the market defines the current market share of natural gas, the other power sources it is competing with and at what relative marginal costs (e.g., coal or nuclear as baseload competition)
- The (future) prevalence of nuclear power defines the role of gas as a baseload or balancing technology in an environment with carbon pricing.
- Age of generation fleet: In a market with old gas plants, natural gas will also tend to be pushed out of the capacity mix earlier.
- The speed of renewable energy expansion driven by declining RES LCOEs
increases the demand for balancing capacities such as gas in the short run; in the long run, RES then outcompete gas.
- The availability and maturity of storage technologies can smooth supply (and demand) fluctuations in most instances in which natural gas would be an alternative.
Three selected countries illustrate the effects from carbon prices
Based on those factors, we can identify three country archetypes that illustrate how country-specific conditions lead to different manifestations of those general trends:
- Increased gas demand with higher carbon prices compared to today due to coal-to-gas switching (e.g., Germany)
- Marginal impact on gas demand with higher carbon prices compared to today due to RES balancing need after coal was replaced (e.g., UK)
- Decreased gas demand with higher carbon prices compared to today due to gas-to-RES switching in countries with a high share of gas (e.g., Mexico)
For our analyses, we compared a base case to three scenarios. For the base case, we froze the current local carbon prices in 2017 for the entire period (referred to as “today”). In three scenarios, carbon prices increase linearly to $30/t, $50/t, or $70/t of CO2e in 2025. As the UK’s carbon price today is already at $33/t, there is no $30/t scenario for the UK.
1. Germany: gas demand increases with higher carbon prices compared to today
In Germany, higher carbon prices would increase the share of natural gas in the generation mix in the next 15 years. Higher carbon prices increase the share of natural gas in the generation mix by replacing coal. Post 2030, the impact of higher carbon prices on the share of gas declines and it can even have a negative impact on gas as it starts to be replaced by RES.
The resulting shift from coal to gas has an immediate effect on emissions: high prices in the short run accelerate the replacement of coal. A carbon price of $30/t in 2025 (roughly the UK level today) would reduce emissions by ~50 Mt/yr CO2e or 22 percent until 2030 and help meet the 2030 target. The quick uptake in the share of gas already with relatively low levels of carbon prices (i.e. 2020–25) underlines the high sensitivity and low-cost threshold to switching from coal to gas. If carbon prices remained at today’s level (and no other measures such as targeted coal capacity retirements are taken), Germany would fail to reach its 2030 targets.
Higher carbon prices of up to $50/t in 2025 would still yield similar marginal abatement effects (an additional ~40 Mt/yr CO2e or ~18 percent reduction in 2030), primarily due to the replacement of more coal by gas (as also reflected in the graph on coal share). Prices higher than USD50/t post 2030 have decreasing marginal abatement, as the highest emitters—lignite and the least efficient coal plants—would already have left the market.
In Germany, the optimal carbon price for gas consumption decreases over time: $30/t to $50/t until 2025 and $30/t or lower post 2030.
2. United Kingdom: higher carbon prices compared to today only have marginal impact on gas demand
In the UK, higher carbon prices (~$33/t) and a reduction in gas prices have already led to a shift from coal to gas. Since 2012, the share of coal in the generation mix dropped—from ~42 percent in 2012 to ~10 percent in 2016
Increasing the carbon price beyond $33/t would only decrease gas consumption by making it less competitive when compared with renewables in combination with storage.
As gas begins to play a more important role in balancing intermittent sources, its consumption decreases and is less and less sensitive to carbon price increases. As an example, a carbon price more than twice as high ($70/t vs $30/t) only reduces emissions by ~5 percent in 2040.
The current carbon price of ~$33/t (EUA
+ GBP18) has already positioned the UK to reach its 2030 emissions target for the power sector of ~32 Mt/yr CO2e
. In 2016, emissions in the power sector already fell by 24 percent compared to 2012 and even by 62 percent when compared to 1990
3. Mexico: gas demand decreases with higher carbon prices compared to today
In Mexico, the share of gas could be negatively affected by a carbon price; a carbon price of USD30/t could reduce the share of gas by nearly 50 percent in 2030.
In the short run (until 2025), a carbon price in Mexico has limited impact on the share of gas (replacing the remaining ~10 percent share of coal in the generation mix).
At the same time, carbon prices would significantly accelerate the decline of gas after 2025; cheaper RES (and later storage) would outprice gas and reduce the share of gas in generation to 20 to 30 percent by 2030. For instance, with a carbon price of $70/t in 2025 and beyond, the short-run marginal cost (SRMC) of existing gas would be ~$52/MWh—well above the LCOE of solar and wind onshore with $33 and $32/MWh, respectively
From a gas consumption perspective, the optimal carbon price would be sufficiently high in the short run to encourage coal-to-gas switching (~$10/t to $20/t) and would then drop to 0.
A carbon price of $30/t would have a significant impact and reduce emissions by 50 percent or ~50 Mt/yr CO2e by 2030. Interestingly, any further increase in the carbon price would yield only disproportionally smaller reductions in annual emissions in 2030. More than doubling the carbon price to USD70/t would only reduce emissions in 2030 by an additional ~10 percent and accelerate the replacement of gas by RES. The effect of such price levels would also no longer justify the costs imposed on the overall economy.
In conclusion, higher carbon prices significantly reduce carbon emissions and generally favor gas to replace coal in the short term (until 2025–30).
Beyond 2025–30, depending on the country-specific conditions, higher carbon prices help RES replace gas, thereby achieving lower marginal abatement for emissions in 2030.
Overall, the effect of carbon prices on gas depends on local conditions, and carbon prices can either:
- Sustainably increase the gas share in the generation mix—especially in places with a high share of coal
- Have little further effect where coal has already been replaced and gas is needed for balancing or
- Significantly accelerate gas replacement by RES
While we have discussed three countries in more detail for this article, we expect many countries to follow similar trends as outlined above, when starting from similar country-specific conditions.
In most cases, already relatively low carbon prices shift generation away from coal and thereby generate substantial emission reductions. Already a carbon price of just $30/t can achieve that, as shown in the UK example.