There's no denying it: Global political will is building behind the vision of a net-zero emission society. Europe wants to become the first net-zero continent and even China has recently jumped on the net-zero bandwagon, targeting 2060.
But we’re going to need a lot more than that. As shown below, the sharp emissions reductions Covid-19 caused in 2020 will need to continue if we want to achieve the 1.5 degrees target. However, climate change tends to slide down the policy agenda during times of socioeconomic hardship, so we can expect emissions to resume their upwards trend when this pandemic finally subsides.
The good news is that clean energy has come a long way over the past two decades. Wind and solar power have seen great cost reductions and are leading the decarbonization charge at present. However, there are two big challenges with a deep decarbonization effort led by variable renewable energy (VRE):
- Wind and solar supply only electricity, which represents just 20% of global final energy consumption today.
- Wind and solar are variable and non-dispatchable, requiring additional technologies to supply energy when there is little wind and sun and others dedicated to consuming excess wind and solar power.
These two challenges have recently rekindled interest in the old idea of a hydrogen economy. Hydrogen is a carbon-free fuel that can decarbonize a sizable fraction of the 80% non-electric final energy consumption while simultaneously balancing VRE.
This is the premise we investigated in our recent paper published open access in the International Journal of Hydrogen Energy. In particular, the paper looked at the issue of capital under-utilization involved in the strategy of balancing wind and solar with flexible hydrogen production.
The Modeled System
To properly evaluate the effects of capital under-utilization, we had to include all the major elements of the integrated electricity-hydrogen system: generation, transmission, and storage. The modeled system (based on Germany) is summarized in the image below:
The following technology options were included:
- Ten different electricity generators: onshore wind, solar PV, pulverized coal and natural gas combined cycle plants with and without CCS, open cycle gas turbine peaker plants, hydrogen combined and open cycle plants, and novel gas switching reforming (GSR) concept.
- Lithium-ion batteries for electricity storage.
- Three clean hydrogen generators: GSR, steam methane reforming (SMR) with CCS, and polymer electrolyte membrane (PEM) electrolysis.
- Two hydrogen storage technologies: cheap salt caverns with slow charge/discharge rates and locational constraints and more expensive storage tanks without such limits and constraints.
- Hydrogen can also be imported in the form of green ammonia that is reconverted to hydrogen in reconversion plants included in the model.
In addition, the costs of the electricity and hydrogen transmission network connecting all these technologies to demand are included in the simulation.
The model objective is to optimize the deployment and hourly dispatch of all these technologies to minimize the cost of the entire system.
Four Scenarios
Our study considered two scenarios where hydrogen could only be produced via electrolysis (Green H2) and two where Blue H2 from natural gas with CCS was also allowed.
- NoCCS: All technologies are available except for power or hydrogen production with CCS. PEM is located close to demand.
- CoLoc: Identical technology availability to the NoCCS scenario, except that PEM is co-located with wind close to cheap salt cavern storage.
- CCS: Identical to the NoCCS scenario, except that conventional power and hydrogen plants with post-combustion CO2 capture technology are also made available for deployment. Only the GSR technology is not available.
- AllTech: Identical to the CCS scenario, except that GSR is also available for deployment. GSR is a novel flexible power and hydrogen production technology designed for the economic integration of higher shares of VRE.
The NoCCS and CoLoc scenarios differ in terms of the placement of the electrolyzers. In NoCCS, electrolyzers are located close to demand, meaning that wind and solar peaks must be transmitted through the costly transmission network to use hydrogen production for balancing VRE.
In CoLoc, electrolyzers are deployed in the north of the country where the wind resource is good, and cheap salt cavern hydrogen storage is a possibility. This avoids the large electricity transmission costs of the NoCCS scenario, but it increases hydrogen transmission costs and restricts electrolyzers to use only wind power from the north of the country.
The Effect of Hydrogen Demand
It is highly uncertain how much hydrogen will be consumed in the clean energy economy of the future. In this study, demand was varied between 0 and 600 TWh/year, which corresponds to 0–33% of current German non-power oil & gas consumption.
Generation Mix and CO2 Emissions
With a CO2 price of €100/ton, the cost-optimal electricity mix looks like this in the four scenarios:
The two Green H2 scenarios (NoCCS and CoLoc) give similar results. In both cases, higher levels of hydrogen demand strongly increase the required electricity generation because of large demand from electrolyzers (PEM).
However, a disappointing finding from this study is that no increase in wind and solar market share is observed as the level of hydrogen demand increases. For this reason, the CO2 emissions intensity of these scenarios stays relatively high.
The reason for this behavior is the large capital under-utilization involved in using electrolyzers to absorb large wind and solar peaks, as will be discussed in more detail in the following section.
In the Blue H2 scenarios, greater hydrogen demand does not increase the required electricity generation because hydrogen is produced via natural gas reforming. In the CCS scenario, there is no connection between hydrogen and power production, so the optimal generation mix remains unchanged with hydrogen demand.
However, the AllTech scenario shows that greater hydrogen demand increases the share of power production from GSR. When there is some demand for hydrogen, GSR can exploit its ability to flexibly produce either power or hydrogen to balance VRE while maintaining a high utilization rate of most of the plant capital. This increases the VRE share in the optimal electricity mix.
Cost Breakdown
As shown below, the Green H2 scenarios turn out to be considerably more expensive than the Blue H2 scenarios at higher levels of hydrogen demand.
The reason for this is that hydrogen produced from electrolysis will always be more expensive than the electricity used to produce it, whereas natural gas can be converted to hydrogen at a significantly lower cost than it can be converted to electricity. Thus, more hydrogen production increases the levelized cost of electricity and hydrogen (LCOEH) in the Green H2 scenarios and reduces it in the Blue H2 scenarios.
The orange bars labeled “Other” amount to a substantial fraction of the total system cost in the Green H2 scenarios. This cost is examined more closely in the figure below.
Clearly, electricity transmission costs are the biggest component of the "Other" costs in the NoCCS scenario. This scenario locates electrolyzers close to demand, requiring large additional transmission capacity to deliver the electricity produced by distant wind and solar farms.
In the CoLoc scenario, these transmission system costs are considerably lower because electrolyzers are co-located with wind farms in the north of the country. However, hydrogen transmission and distribution (T&D) costs are higher in this scenario because hydrogen must be transmitted from the north throughout the entire country.
The CCS scenario requires only mild investments in H2 T&D infrastructure because hydrogen is generated according to market needs close to demand centers. These costs are significantly higher in the AllTech scenario because flexible power production from GSR implies a more intermittent hydrogen production profile, requiring more transmission and storage capacity to handle the intermittent hydrogen fluxes.
The Effect of CO2 Prices
As illustrated in the previous section, the Green H2 scenarios still produced considerable CO2 emissions, even with a CO2 price of €100/ton. Achieving deep decarbonization will require higher CO2 prices.
Generation Mix and CO2 Emissions
The graph below illustrates the effect of higher CO2 prices on the optimal electricity mix and CO2 emissions intensity. Hydrogen demand is set to 400 TWh/year in all cases.
Clearly, higher CO2 prices substantially reduce the CO2 emissions in the Green H2 cases by displacing unabated natural gas-fired power generation with wind and solar.
In the Blue H2 scenarios, an increase in CO2 price from 50 to 100 €/ton had a large effect by incentivizing CCS in the power sector. Beyond this point, further increases in CO2 price have only a small effect because CO2 emissions are already very low at €100/ton. Most notably, higher CO2 prices incentivize more GSR and VRE in the AllTech scenario.
Cost Breakdown
As shown below, the reduction in CO2 emissions in the Green H2 scenarios comes at a cost. Given their generally lower emissions, the cost of the Blue H2 scenarios is less sensitive to increased CO2 prices.
In the NoCCS scenario, €200/ton is enough to allow the system to transition to using electrolyzers instead of NGCC plants as the primary mechanism for balancing VRE. This is reflected in the considerable reduction in "Unabated" power production costs and the increase in "Other" costs when the CO2 price is increased from 150 to 200 €/ton. Other costs increase sharply because this strategy requires large transmission network overbuilds to transmit VRE peaks to electrolyzers.
The CoLoc scenario shows a smoother trend. Here, increased CO2 prices also incentivize more VRE balancing via electrolysis instead of NGCC power plants. This also results in a steady increase in "Other" costs due to the lower utilization rate of electrolyzers, H2 transmission pipelines, and H2 storage infrastructure when handling increasingly pronounced peaks of intermittent hydrogen production.
This scenario is also highly dependent on the availability of cheap salt cavern hydrogen storage close to the co-located wind and electrolyzer capacity. Such capacity is scarce around Europe, and its exploitation could face considerable public resistance. If more expensive tank storage must be used, system costs increase to the level of the NoCCS scenario, and 40% of hydrogen demand must be imported.
Only minor effects are observed in the Blue H2 scenarios, mainly the aforementioned transition from unabated power plants to CCS power plants when the CO2 price is increased from 50 to 100 €/ton.
Conclusions
The main conclusion from this study is that, although hydrogen can be used to integrate higher shares of wind and solar, this strategy brings considerable costs due to capital under-utilization.
- When electrolyzers are co-located with demand, expensive transmission network expansions are required to transmit wind and solar production peaks to electrolyzers.
- When electrolyzers are co-located with wind power, the low utilization of electrolyzers and the large hydrogen transmission and storage capacity required to handle intermittent hydrogen fluxes inflate system costs.
- When conventional CCS power plants are deployed, the model chooses to operate these plants under baseload conditions to maximize the utilization of expensive CCS infrastructure, limiting VRE deployment.
- Flexible power and hydrogen production from GSR can integrate more wind and solar, but the associated intermittent hydrogen production increases hydrogen transmission and storage costs, reducing the positive impact of this novel process.
Such a whole-system perspective is critical for optimizing the rollout of the energy transition. Given the high level of technology interdependence involved in such integrated electricity-hydrogen systems, careful planning is required to minimize costs and complexity. Blue hydrogen has an important role to play in this regard and should not be dismissed from the policy agenda.