Hawaiʻi’s LNG Business Case Was Overly Optimistic & Built On A Broken Spreadsheet

Support CleanTechnica’s work through a Substack subscription or on Stripe.

The revelation that the spreadsheet behind Hawaiʻi’s headline LNG savings case appears not to include the cost of the LNG itself is the kind of finding that changes the center of gravity of an entire policy debate. The Sierra Club of Hawaiʻi press release on March 12, built around comments by former University of Hawaiʻi engineering professor and Energy Innovations Director of Global Policy Research Matthias Fripp at a House Energy and Environmental Protection Committee briefing, said that the Energy Office’s widely cited Scenario 3A graph showing roughly $700 million to $800 million in net savings failed to incorporate nearly $900 million in LNG fuel costs.

Fripp’s account was even more damaging than the number. He said he multiplied assumed LNG prices in the spreadsheet by 100 and the graph did not change, which is exactly the kind of stress test that reveals a disconnected formula rather than a tough modeling disagreement. He validated his findings with several others. If that description of the workbook is correct, then the most publicized economic case for LNG on Oʻahu did not just wobble. It inverted.

That matters because the Hawaiʻi State Energy Office’s January 2025 Alternative Fuels, Repowering, and Energy Transition Study was not sold as a narrow engineering memo. It was presented as part of a larger strategy to replace petroleum fuels, attract investment, enhance resilience, improve affordability, and accelerate renewable adoption. The study sits in a context where Hawaiʻi has some of the highest electricity prices in the United States, where Oʻahu remains highly dependent on imported oil, and where the post-Maui wildfire financial and political environment made grid reliability, capital attraction, and institutional confidence even more salient. The state was trying to solve real problems. That is exactly why the arithmetic, assumptions, and delivery timelines deserve hard scrutiny instead of a soft political reading.

The official LNG case, at least in its public form, had a straightforward narrative. Oʻahu would keep building solar, wind, batteries, and other renewables, but because that buildout might not happen fast enough to retire thermal plants and keep the lights on, the state should import LNG as a transitional fuel. The main physical concept in the study was an offshore floating storage and regasification unit near Barbers Point, a subsea pipeline to shore, onshore pipelines to key plants, a conversion of Kalaeloa Partners, and new combined cycle capacity at Barbers Point and later Kahe. The study assumed Phase 1 service by 2030 and Phase 2 by 2035, and framed LNG as a bridge that could later lead to hydrogen or ammonia based power in 2045. This was not merely a fuel swap. It was a multistage import terminal, pipeline, and power generation program tied to the state’s 2045 clean energy ambition.

The players around that narrative also matter because the proposal did not emerge from a vacuum. The study itself builds on work by HSEO, HDR, Facts Global Energy, NREL, Hawaiian Electric, and others. Governor Josh Green’s administration later deepened the political signal by signing a strategic partnering agreement with JERA in October 2025, with both the governor’s office and JERA explicitly linking the partnership to the January 2025 study and to fuel diversity and decarbonization pathways. Around the same time, the Coalition for Hawaiʻi’s Energy Future emerged as a public advocacy network with more than 50 local partners spanning business groups, nonprofits, hospitality, construction, agriculture, and logistics, all pushing a message of affordability, reliability, lower emissions, and private investment. This is important because once a proposal gains institutional sponsors, international commercial interest, and local coalition support, optimistic scenarios can begin to feel like plans instead of contingencies.

What the public messaging left understated is that the economics in the HSEO study were never robust across the full range of scenarios. The study’s own structure turns on whether LNG infrastructure can later be reused for hydrogen, what fuels LNG is assumed to displace, and whether the renewable portfolio standard path is met on time. In Alternative 1A, the benefits of an interim LNG transition exceed the costs by about $150 million in net present value, or only $10.2/MWh, and even that modest upside depends on LNG prices rising by no more than 10%, low sulfur fuel oil prices falling by no more than 5%, and capital costs rising by no more than 20%. That is not a broad margin for error. It is a narrow corridor.

The more widely promoted case was Alternative 3A, which HSEO itself describes as a more optimistic future scenario. In that case, LNG does not merely displace LSFO. Capacity expansion runs imply that it displaces a mix of LSFO, some utility scale solar, and biodiesel. That matters because the more expensive the displaced fuel mix, the better LNG looks. Under that optimistic stack, HSEO reported an NPV benefit of about $867 million, levelized savings of $59/MWh, and residential savings of about 15.2%, or roughly $352 per year. It also said the evaluation included all fuel cost savings of biodiesel, some solar, and some LSFO. In other words, the headline case already depended on a more favorable displaced fuel mix than the simpler bridge fuel story suggested. It was a compound best case before the spreadsheet issue surfaced.

That imported fossil fuels were supposed to displace some utility solar and that was considered a positive part of the plan is also astounding.

The darker half of the public record sits in the scenarios that received less attention. HSEO states that Alternative 2 explores a transition to an undefined non-hydrogen renewable fuel source that does not allow reuse of most LNG infrastructure. In that branch, fuel savings alone are not enough to generate cost savings for ratepayers, and HSEO says Alternatives 2B and 2C, while more favorable than 2A, still did not result in cost savings. That is the hinge of the entire exercise. If the hydrogen reuse story weakens, the LNG economics collapse. The study itself makes that explicit, noting that reuse of infrastructure constructed for a methane gas transition strongly impacts the results. The bridge is only attractive if the destination turns up on time and at acceptable cost.

That is what makes the alleged spreadsheet omission so consequential. A model that already depended on optimistic displaced fuel assumptions, hydrogen asset reuse, and tight cost thresholds appears, based on Fripp’s public description, to have left out roughly $900 million in LNG fuel cost from the scenario that was held up as the strongest affordability case. The Sierra Club press release said that once the error and other omissions were considered, the supposed $700 million to $800 million in net savings could become $300 million to $400 million or more in net costs. Even if that range shifts with fuller disclosure, the directional point is clear. The public headline was built on a stack of favorable assumptions, and removing one major accounting error does not leave a small positive residue. It appears to flip the sign.

This is where Bent Flyvbjerg’s reference class forecasting, accessibly described in his 2023 book with Dan Gardner, How Big Things Get Done, becomes useful. The inside view asks whether Hawaiʻi’s specific project team, procurement strategy, marine conditions, and permitting path might succeed. The outside view asks what happens to projects like this in the real world. For megaprojects, that outside view almost always performs better. Research on major projects reports that reference class forecasting cut average cost overrun from about 50% to 5% where it was applied. The logic is not mystical. It strips away project specific optimism and forces planners to confront actual distributions of cost and schedule performance in comparable classes of projects. Hawaiʻi’s LNG plan deserves exactly that treatment because it was advanced on the basis of a bespoke case with ambitious timing and unusually sensitive economics.

The relevant reference class is not a generic public works project. It is a hybrid of LNG import terminals, floating storage and regasification units (FSRU), marine energy infrastructure, subsea pipelines, island logistics, and linked thermal repowering. The closest public comparators are LNG megaprojects and FSRU import terminals. EY’s survey of 365 oil and gas megaprojects found that 64% had cost overruns and 73% had schedule delays. For LNG megaprojects specifically, 67% had cost overruns, 68% had schedule delays, and average budget overruns reached about 70%. For North American oil and gas megaprojects, 58% had cost overruns, 55% had schedule delays, and average overruns were about 51%. None of those numbers guarantee failure in Hawaiʻi. All of them say that the default expectation for a project of this class should be higher cost and later delivery than the sponsor case.

Now compare that reference class to the Hawaiian import facility numbers. HSEO’s report and appendix frame the import side at roughly $412 million for the FSRU, buoy system, and subsea pipeline, plus about $12 million for onshore pipeline links to key plants. Facts Global Energy’s public deck framed the import system with a roughly $400 million charter scenario and a roughly $700 million purchase scenario for the LNG import system, excluding new power generation capex. Apply a 51% North American oil and gas uplift to the $412 million core import package and it rises to about $622 million. Apply a 70% LNG megaproject uplift and it rises to about $700 million. That is striking because a plain outside view correction takes the study’s low import estimate and lands it right on top of the higher scenario before discussing financing costs, marine claims, or delay effects.

The schedule side is just as telling. Studies of LNG infrastructure show that onshore LNG terminals are usually driven by tank construction and typically take 36 to 40 months. New build FSRUs typically take 27 to 36 months, while conversions take about 18 to 24 months. The real schedule advantage only appears when an appropriate vessel is readily available or already under speculative construction. Hawaiʻi is not proposing merely to secure an FSRU. It is proposing, by 2030, to have an offshore mooring and regasification setup, a subsea pipeline, shoreline interface, onshore pipelines, a plant conversion at Kalaeloa, and a new combined cycle build at Barbers Point, all under Hawaiʻi’s marine, land use, and permitting conditions. The HSEO report itself warns of construction, supply chain, funding, financing, and permitting risks. Analysts reviewing the proposal have summarized the project as highly sensitive to fuel price dynamics, infrastructure timing, post 2045 firm power assumptions, and system level interactions. That is not the language of a schedule with slack.

If one takes the outside view seriously, the chance of meeting the project’s required schedule is not high. A normal new build FSRU timeline of 27 to 36 months already absorbs most of the 2027 to 2030 window assumed in the HSEO study before adding the subsea and onshore pipeline work and linked power plant work. The megaproject data suggest that delays are more common than on time delivery. It is not enough to say that Hawaiʻi could do better than the average. The positive economics in the study require Hawaiʻi to do better than the average while also keeping capital costs contained and preserving the 2045 hydrogen reuse story. Once those conditions are stacked together, the probability of all of them holding at once drops quickly.

The hydrogen piece is where the proposal becomes even less grounded. HSEO’s own report says the economic evaluation considered whether LNG infrastructure could be reused for a future renewable energy solution such as hydrogen. It also provides a conceptual 2045 system in which the FSRU and subsea pipeline are no longer in service, onshore pipelines are converted for hydrogen service, and new ammonia receiving and unloading infrastructure is added. That is already an admission that the LNG import assets are not simply transformed into a hydrogen system. Much of the offshore import side is stranded by design, and major new ammonia and hydrogen handling infrastructure still has to be built. The study describes this as an early stage assessment and says cost reduction, scalability, and infrastructure advances are still required.

Against that, the hydrogen cost literature I have published over the past few years points in a different direction. In my analysis of electrolysis cost projections early in 2025, I argued that green hydrogen at the plant gate in 2050 is more likely to be in the $6 to $8/kg range under realistic assumptions. In my analysis of European hydrogen infrastructure economics earlier this year, I argued that delivered green hydrogen in Europe remains in the €8/kg to €12/kg range under realistic assumptions. The logic is simple. Hydrogen is capital heavy, electricity intensive, storage intensive, and transport sensitive. Once those layers are included, the cheap hydrogen story collapses. At 33.3 kWh/kg of lower heating value, $6/kg hydrogen already means $180/MWh of fuel energy. At $8/kg it becomes $240/MWh. Run that through a 55% efficient combined cycle plant and the fuel cost alone becomes roughly $327/MWh to $436/MWh of electricity before capital recovery, storage, backup fuel, and operations.

Shipping hydrogen does not rescue the economics. In my earlier work on liquid hydrogen shipping economics, I found that liquid hydrogen would be at least 5 times as expensive as LNG per delivered unit of energy in a best case scenario. The result comes from the combination of much lower volumetric energy density, high liquefaction energy demand, and the same ocean transport cost structure spread over far less delivered energy. Even using optimistic assumptions, liquefied hydrogen remained structurally expensive. I made the same point again in later work, noting that even truck delivered hydrogen in best case European cases was already around $8/kg before manufacturing cost and markup. If Hawaiʻi were to imagine imported hydrogen as a routine firm power fuel, it would be starting from a chain whose transport economics are already worse than the fuel it is trying to replace.

Imported ammonia, the more realistic carrier proposed for the Hawaiian endgame, is not much better once the math is carried through. In my 2024 analysis of ammonia as an energy carrier, I used a delivered ammonia case of about $1,000 per ton and showed that direct ammonia combustion for power works out to roughly $900/MWh after accounting for the hydrogen fraction of ammonia, conversion efficiency, storage, and handling. Cracking ammonia back to hydrogen and running it through a fuel cell improved the thermodynamics but still landed around $420/MWh before adding realistic risk margins for fuel cells and the rest of the chain. Those are wholesale generation economics, not retail prices. They are nowhere near the range that an island grid struggling with affordability should treat as a likely mainstream end state.

This isn’t just my assessment. Recent scrutiny by European oversight bodies also reinforces caution around hydrogen economics. France’s Court of Auditors concluded in 2024 that the country’s hydrogen strategy carries substantial fiscal risk because production costs remain far above those of existing energy carriers, with public subsidies likely to cover large portions of the gap for years. Germany’s Federal Court of Auditors reached a similar conclusion, warning that the government’s hydrogen strategy rests on optimistic assumptions about future supply and import costs while committing billions in infrastructure spending before demand is certain. Germany’s Council of Economic Experts has likewise noted that hydrogen is likely to remain a scarce and expensive energy carrier best reserved for sectors such as steel, chemicals, and certain industrial processes rather than widespread use in power generation. Together these assessments from national audit institutions and economic advisers point to the same basic conclusion. Hydrogen may have important niche roles in deep decarbonization, but treating it as a broad replacement fuel for electricity systems carries significant economic risk.

That leaves the LNG bridge resting on a destination fuel pathway that is weakly specified in the state study and poorly aligned with the outside cost literature. The positive scenarios in HSEO’s work require meaningful reuse of LNG infrastructure for hydrogen, but the state’s own infrastructure concept shows the FSRU and subsea pipeline disappearing by 2045 and new ammonia infrastructure appearing in their place. The less optimistic scenarios already show that without reuse the LNG economics turn negative. My published work on hydrogen production, shipping, and ammonia to power suggests that the post 2045 hydrogen pathway is unlikely to be cost competitive for routine generation. That means the proposal depends on a future asset salvage story that is not only under modeled in the public documents but economically doubtful on first principles.

The organizations pushing LNG in Hawaiʻi are not hard to identify. HSEO produced the state study and gave it institutional legitimacy. Governor Josh Green’s administration amplified it and formalized collaboration with JERA, one of the world’s largest LNG buyers. Facts Global Energy produced supporting fuel and economics work arguing LNG could save billions under favorable assumptions. The Coalition for Hawaiʻi’s Energy Future built a local political coalition around affordability, reliability, and transition messaging. None of this means the participants are acting in bad faith. It means the proposal developed a political and commercial ecosystem quickly, with each actor seeing something different in it. State officials saw reliability and affordability. International LNG interests saw a foothold in an island market. Local businesses saw a promise of lower rates. Critics saw cost, climate, and stranded asset risk.

Put together, the standard that should be applied in Hawaiʻi becomes clear. A project of this scale should be evaluated using full auditable fuel cost accounting rather than simplified public graphics. It should be stress tested against the less favorable scenarios in its own appendix rather than relying on the optimistic one. It should be adjusted using reference class forecasting for both capital cost and schedule because LNG import infrastructure and associated repowering do not live in a world where sponsor estimates are usually correct. And if hydrogen reuse is central to the favorable economics, then the state should publish an explicit delivered cost case for hydrogen and ammonia power that can be compared directly with other options.

What began as a spreadsheet problem ends up looking more like a systems problem. The most publicized LNG case in Hawaiʻi appears to have relied on a scenario that was already more optimistic than the plain language sales pitch suggested. It appears to have required aggressive infrastructure delivery on a project class with a long history of overruns and delays. It appears to have depended on future hydrogen reuse to keep the economics positive even though the hydrogen endgame remains weak on public numbers and poor on outside view economics. And if the reported LNG fuel cost omission is valid, the headline savings case was not merely optimistic. It was arithmetically unsound. Hawaiʻi has real reliability and affordability challenges, but solving them requires plans that survive contact with both the spreadsheet and the reference class.