From carbon neutral to carbon negative: a theoretical bioenergy and CO2 removal retrofit at Ngāwhā geothermal power station

Titus, Karan; Dempsey, David; Peer, Rebecca; Hanik, Fabian

doi:10.1080/03036758.2024.2385807

We modelled a biomass boiler retrofit that adds 1 MWe to Station-1 at the Ngāwhā geothermal power station. Dissolving the biomass CO2 in reinjected fluid could remove up to 15,900 tonnes of CO2 a year. The design that also sells its surplus CO2 had the lowest costs in our model.

The problem

New Zealand generates at least 80% of its electricity from renewable sources (MBIE, 2021), but almost half of its greenhouse gas emissions come from agriculture (McDowell et al., 2024). More renewable electricity cannot offset these emissions. As of 2023, the country's only negative emissions strategy was afforestation.

Geothermal fluid contains dissolved gases such as CO2 (McLean et al., 2020), which power plants have traditionally vented. Ngāwhā power station, in the Far North, dissolves these gases back into its reinjected fluid, and in 2023 it became New Zealand's first carbon neutral geothermal plant. Burning biomass releases biogenic CO2, which is treated as carbon neutral (Lehtveer and Emanuelsson, 2021). Storing that CO2 underground could make the plant carbon negative. In this study, we asked what such a retrofit of Ngāwhā would cost and how much CO2 it could remove.

What we did

We studied Station-1, the two oldest binary units at Ngāwhā. In our retrofit, a biomass boiler fuelled by Far North forestry residue heats the fluid before the separator. The extra steam lets the binary units run their working fluid at a higher pressure and generate more power.

We compared three configurations, each sized to add 1 MWe (megawatt of electricity) of net output because of limits on the existing generator. In Configuration 1, all biogenic CO2 is vented. In Configuration 2 (geothermal-BECCS), as much as possible is dissolved in the reinjected fluid and the rest is vented. In Configuration 3 (geothermal-BECCUS), the excess is sold for uses such as food production. Configurations 2 and 3 burn biomass in pure oxygen to give a flue gas of nearly pure CO2. Net output deducts the 0.8 MWe (Configuration 2) or 1.2 MWe (Configuration 3) used by the oxygen plant and CO2 compressor.

We capped the total CO2 injected so that it would dissolve completely at 20 bar and 95 °C. We chose a pressure 2 bar below the injection pressure so that small changes in fluid temperature would not bring CO2 out of solution. Allowing for the 0.5% by mass of gas already present, the fluid can dissolve a further 1.8 tonnes of biogenic CO2 an hour.

We used our thermodynamic systems model (Titus et al., 2023a, 2023b) with a 15-year retrofit life, a 90% capacity factor and an 8% discount rate. The levelised cost of new electricity (LCNE) is the retrofit cost, less revenue from CO2 removal and sales, per MWh of new electricity. The levelised cost of removed carbon (LCRC) is the retrofit cost, less revenue from electricity and CO2 sales, per tonne of CO2 removed. We sampled five uncertain inputs 1000 times and report medians. Costs are in US dollars. Our base case prices CO2 at $50 a tonne, from the 2022 cost containment trigger of New Zealand's emissions trading scheme (ETS) (Reddy, 2023), and electricity at $60 per MWh. We assumed that CO2 is sold at the ETS price.

What we found

We find that Configurations 2 and 3 each remove up to 15.9 kt (thousand tonnes) of CO2 a year, the most the reinjected fluid can dissolve. This would offset roughly twice the CO2 that a 1 MWe coal power plant emits (EIA, 2021). In the base case, CO2 removal would earn $0.79 million a year, CO2 sales in Configuration 3 a further $1.8 million, and the new electricity $0.47 million. Configuration 3 burns 26.5 kt of biomass a year. Using supply data for Northland (MPI, 2020; Nurek et al., 2019), we estimate that this is about 6% of the region's long-term supply of chip and hog quality forestry residue.

The median LCNE was $194, $267 and $137 per MWh for Configurations 1 to 3, against the $60 we assumed the electricity sells for. Configuration 3 had the lowest LCNE. This is mainly because its CO2 sales, about $206 per MWh, exceeded its capital cost, about $196 per MWh. The median LCRC was $164 a tonne for Configuration 2 and $93 for Configuration 3. Removal costing more than $200 a tonne may be considered uncompetitive with other removal technologies (Friedmann et al., 2020). Both medians were below that threshold. Only Configuration 3 was below $100 a tonne, a cost we suggest is feasible.

Configuration 3 had a lower LCNE than Configuration 2 above about $20 a tonne, and than Configuration 1 above about $40. Above about $70, its LCNE was negative. At those prices, CO2 revenue exceeded the whole cost of the new electricity. The median LCRC of Configuration 2 remained below $200 a tonne even at an electricity price of zero. However, Configuration 2 was unlikely to reach $100 a tonne at electricity prices up to $120 per MWh. Configuration 3 was cheaper at every electricity price we tested.

Limitations

These results assume that CO2 removal will eventually earn revenue through the ETS or a similar future mechanism. Configuration 3 also assumes buyers for its 36.1 kt of surplus CO2 a year. With no such plant operating as of 2024, the uncertain inputs came from ranges in the literature. We sampled them independently.

The existing separator must be able to handle the extra steam. The generator's spare capacity, about 2 MW before the oxygen plant and compressor loads, may also limit the extra power. How quickly the dissolved CO2 turns into minerals at Ngāwhā is unknown. We recommend a pilot demonstration as the next step.

Glossary

MWe is megawatts of electricity. MWh is megawatt-hours of energy.

kt is thousands of tonnes.

Biogenic CO2 comes from burning plant material, which took that carbon from the atmosphere as it grew.

A carbon neutral plant adds no net CO2 to the atmosphere. A carbon negative plant removes more than it emits.

Reinjection returns used geothermal fluid to the reservoir through wells.

A binary plant passes heat from geothermal fluid to a second working fluid in a closed loop that drives the turbine.

A separator splits geothermal fluid into steam and liquid.

Flue gas is the exhaust gas from a boiler.

Chip and hog quality residues are forestry offcuts that can be chipped or shredded into boiler fuel.

BECCS is bioenergy with carbon capture and storage. BECCUS adds utilisation, the sale and use of captured CO2.

The capacity factor is the share of the year a plant runs at full output.

The discount rate reduces the value of future costs and revenues relative to present ones.

A levelised cost spreads a project's lifetime costs, less its other revenue, over its lifetime output.

The median is the middle value of the sampled results.

The emissions trading scheme (ETS) is New Zealand's market for emission units. Its cost containment trigger is the highest price before reserve units are released for sale.

Abstract

For countries like New Zealand with high renewable electricity generation (≥80%) but large emissions per capita, traditional decarbonisation methods are limited or costly. However, carbon dioxide removal (CDR) technologies can drive multiple sectors of the economy across the net‐zero barrier through negative CO2 emissions. We argue that the key to scaling up CDR begins with the preponderance of New Zealand's geothermal and biomass resources. New Zealand has a proud and innovative history with geothermal energy, currently producing ∼20% of the country's electricity. Modern advances in geothermal energy have demonstrated that it is peerless amongst renewable energy‐sources in the ability to facilitate onsite CDR by repurposing existing wells. We examine a theoretical bioenergy retrofit at the Ngāwhā geothermal plant to increase capacity by 1 MWe, with as much biogenic CO2 as permissible dissolved in reinjectate. Forestry residues are the feedstock of choice due to their abundance in the Far North. We show that up to 15.9 ktCO2 /year can be removed effectively from the atmosphere. Only 6% of the Far North's forestry residues are required to achieve this. Under 2024s emissions trading scheme (ETS), the annual revenue of CDR ($0.79 million) could exceed that of new electricity ($0.47 million).