The Petra Nova CO2 capture project – why not open it up?

The recent sale by NRG of their 50% stake in the Petra Nova CO₂ capture facility for a paltry $3.6M, less than 1% of its construction cost, suggests that the economics of CO₂ capture for enhanced oil recovery (EOR) at this site are not very attractive.  While actual details of the plant performance are lacking, even in the official DOE project report, capture cost data published by the GCCSI (see above) indicated that solvent management costs alone were of the order of $10 per tonne of CO₂ captured.  Bearing in mind that the value of CO₂ for EOR implied in the 45Q tax credit mechanism is $15 per tonne (the difference between the $50/tCO₂ for saline aquifer storage and $35/tCO₂ for EOR), it appears that even Petra Nova’s full operating cost may not have been recoverable, let alone its capital costs.  In addition, some maintenance expenditure may have been found to be necessary after a shut-down in 2020, when oil prices were low.  The new sole owners, formerly 50% stakeholders, JX Nippon have apparently stated that “no clear timeline for the restart has been set, but we aim to be ready for resuming the CO₂ capturing facility in the second quarter next year (i.e. 2023) when the power plant is due to be back online,” also noting that the power plant suffered a fire in May this year. 

How did the Petra Nova project get here?  A possible answer is that the promise of low energy consumption for releasing captured CO₂ was allowed to trump all other considerations, in particular the ability to reclaim clean amines for reuse from the dirty mixture with flue gas contaminants and degradation products that amine solvents can turn into when capturing on coal plants for extended periods.  To achieve low energy consumption a blend of two or more amines is typically used, a high-capacity, low-energy component with relatively low intrinsic reactivity plus an accelerator amine to give acceptable overall reaction rates between the solvent and the flue gas as they meet in a contactor tower many tens of metres tall.  But, while these more sophisticated blends do achieve reduced energy compared to the proven, industry-standard monoethanolamine (MEA) solutions, they are intrinsically much more difficult to reclaim.  As a single component, MEA can be boiled out of the circulating solvent in a simple ‘kettle’ reboiler with no more difficulty than distilling whisky, and the energy used can be recovered as well when the MEA and water vapour are vented into the main solvent regeneration unit.  In contrast, for a blend of two or more components with inherently different boiling points, both recovering all of the original amines and also rejecting unwanted degradation products that co-distil becomes much more challenging.  This difficulty in reclaiming, plus intrinsically-higher costs for more complex amines, may explain the $10/tCO₂ cost reported by the GCCSI.  In addition, although apparently not discussed in connection with Petra Nova, low energy solvents, in particular the accelerator component if it is a secondary amine such as piperazine, may often readily react with NOx in the flue gases to make stable nitrosamine compounds that are highly carcinogenic, something that primary amines like MEA cannot do.

So, the solution to perhaps as much as a third of Petra Nova’s operating cost problem could be as easy as switching to a fully reclaimable, and low cost, solvent like MEA.  In a recent post-combustion capture FEED study using MEA it was estimated that solvent management (make-up and waste disposal) would cost $3.65/tCO₂, although on a GT flue gas with a high oxygen content rather than on a coal plant flue gas with lower oxygen but typically higher dust and SOx content.  Some extra operating energy costs would probably be incurred too with MEA, but the GCCSI figures suggest that energy has not been the dominant cost component and energy costs may become even less important as fossil power plant, and hence capture unit, load factors decrease with increasing renewable generation input to the Texas grid.  It is very likely, too, that using MEA in a post-combustion capture plant would be feasible for the flue gas at this site because, when the project was originally developed, it was proposed to use Fluor’s MEA-based Econamine FG Plus solvent.   So, perhaps if the existing plant were to be transferred to another operator at its apparent valuation of $7.2M (based on the recent sale), it might be operated successfully using MEA, with the added benefits of no risk of direct nitrosamine formation and also, because MEA is an ‘open-art’ solvent, the opportunity for full disclosure of all aspects of plant performance.  Of course, JX Nippon may still wish to continue with a proprietary solvent; in this case the MEA solvent option might still be useful as a reserve in case problems are encountered in the future.

GCCSI analysis of levelised cost of capture for BD3, Petra Nova and a proposed retrofit plant at Shand (data from GCCSI, Global Status of CCS Report, 2019) ‘Variable O&M’ costs are likely to be predominantly for solvent management and replacement. Results based on 8% discount rate, 30 years’ project life, 2.5 years’ construction time, capacity factor of 85%.  Cost data are normalised to 2017 values. Stated expected accuracy range: Boundary Dam and Petra Nova: -10% to +15%, Shand: -25% to +40%.

Jon Gibbins, UKCCSRC Director