A longtime Japanese LNG industry engineer is raising fundamental concerns about the schematics behind major portions of the $43 billion Alaska LNG Project.
A semiretired process engineer from Tokyo, Keiji Akiyama said in multiple interviews that he believes the Alaska Gasline Development Corp. is on a path to repeating the mistakes made in numerous other large LNG projects across the globe without wholesale changes to the design of the LNG plant with capacity of 20 million tons per year planned for the terminus of the project in Nikiski.
Akiyama, most recently a senior LNG technical adviser for INPEX Corp., Japan's largest oil and gas company, according to his resume, said he began following the Alaska LNG Project after an industry friend asked him to review the design of the small Kenai LNG plant when ConocoPhillips put it up for sale in late 2016.
That plant has since been sold to Andeavor Corp., which also owns the Nikiski-area oil refinery.
Akiyama, with more than 40 years in the industry, said he was peripherally involved in Alaska's early gasline ventures 20-plus years ago but hadn't tracked more recent efforts to monetize the North Slope's huge gas resources.
Since, though, he has been involved primarily as a contractor in many LNG developments throughout the Middle East, Southeast Asia, Australia, and most recently Russia's Arctic Yamal project.
"I saw that it's a good opportunity to look into some detail. I know how people tend to think about energy projects from my experience; so I have focused on how most people will look at this project with a primarily commercial view or even, as you say, a political view, but not technical view, no," he told the Journal.
"That is always an issue because from my experiences if plant doesn't work as expected it's extremely painful, extremely painful and sometimes it will kill the country because of the size of the project."
Akiyama prepared a 26-slide PowerPoint presentation for AGDC and added that he has come to understand the bind declining oil revenues have put the State of Alaska in how and important the Alaska LNG Project could be for the state.
"If government will see an unexpected result, it's really a disaster for everybody and if I can do something I should do it as expert of technical aspect," he said.
Plot points
The issues he insists he has identified largely relate to the layout of the massive Alaska LNG plant, which would be located on a parcel of more than 600 acres.
A simple overhead view, blueprint-style illustration of one of the three LNG trains planned for the plant, which Akiyama refers to as a "plot plan," shows rows and rows of heat exchanger fans running east-west across much of the structure.
Akiyama is concerned the heat exchangers are situated too close to the compressors that intake air to start the chilling process. That's because the fans release air hot with waste heat from the liquefaction and if the compressors intake the hot air it could severely hamper the efficiency of the plant, requiring more fuel gas to produce the LNG product.
AGDC leaders have long touted the plant efficiency advantage Alaska maintains over LNG projects in warmer climates for the simple reason the air used in the system won't need to be cooled quite as much.
Akiyama compares it to multiple air conditioning units, acknowledging Alaskans might not be as familiar as most people with the appliances.
"If you have a so-called outside unit in your garden, you have a lot of warm air coming out. Air conditioning machine is exactly the same as LNG, same thing," he said. "Finally, all warm air will come up from air-cooled heat exchangers, particularly propane condenser. Propane condenser will release a lot of warm air and most people will never install a lot of air conditioning machines in your garden too close to each other."
The documents he referenced, at one point labeled confidential, are part of AGDC's public filings with the Federal Energy Regulatory Commission to support the project's environmental impact statement application.
Nikiski's summer winds could add to the issue of hot air recirculation, according to Akiyama.
He said most plants are designed with prevailing wind in mind, but the focus is usually on the average prevailing wind direction over an entire year.
Cold Alaska winter winds need not be accounted for, Akiyama said, but Nikiski's common summer wind blowing up Cook Inlet from the south-southwest could compound the impact of the hot air recirculation by blowing the hot air into the compressors.
Hot air recirculation has hampered production at a number of LNG plants worldwide, according to Akiyama.
The West Australian, a newspaper in Perth, has reported that Chevron's massive Gorgon project on an island off the country's coast — pegged at $37 billion when it was started in 2009 and finished for $54 billion — has seen its expected production cut by 13 percent because of the problem. It could cost the company up to $500 million per year and challenge Chevron's ability to meet its contracts, according to a report from the paper.
"I got in a big fight with them," said Akiyama, who worked as a contractor on Gorgon. "'Don't do it; don't do it; let's think, stop,' but they didn't listen. That is the very first LNG project for Chevron."
He said further that hot air recirculation has caused the North West Shelf LNG project, also in Australia, to completely shut down at times.
Akiyama said he informed Masatoshi Shiratori, AGDC's marketing representative in Tokyo, of his concerns, but he wasn't aware if they had been relayed to officials in Anchorage.
AGDC spokesman Jesse Carlstrom wrote in an emailed response to questions that it does not appear Shiratori discussed Akiyama's conclusions with anyone in Anchorage.
AGDC engineers do not believe hot air recirculation will be an issue, according to Carlstrom.
"The hot air flow was checked with seasonal wind speed and direction considered, and additional heat exchanger capacity was added as appropriate for the project economics," he wrote. "The seasonal wind directions in Nikiski are well-known, and this was factored into the simulation."
At a more basic level, Akiyama also questioned whether the layout of the plant allows enough room for maintenance cranes to maneuver amongst the fixtures.
While the plot plan illustrations Akiyama has studied are relatively simple sketches, he contends much can be gleaned from them and they are among the most important documents for a large project.
"So although the plot plan once completed is merely a drawing, we have to think and imagine from various aspects. So it takes a long time to train people to develop the plot plan. Looks like a very simple drawing but actually it is not. A lot of consideration for process, safety, construction, everything — of course, money," he said. "So once I have a plot plan on my desk I can tell it is more or less ok or a very bad job or so-so. And although they didn't disclose a lot of technical documents, fortunately or unfortunately, they have disclosed the most important document, site layout."
Akiyama believes the issues are fundamental enough to warrant redrafting and resubmission of the documents to FERC, which he notes could set the project back by a year or more. He suggested extending the LNG trains to farther to the east into what is now planned as a lay down area.
AGDC President Keith Meyer is pushing hard for making a final investment decision on the Alaska LNG Project in late 2019 or early 2020 — which is later than he initially targeted after the state took over the project in 2017 — or as soon as FERC issues a favorable record of decision.
The documents submitted to FERC were largely compiled when ExxonMobil led the project and then transferred to AGDC at the start of 2017 and Akiyama said he doesn't understand why a company with significant experience in LNG would put together what he sees as significantly flawed work.
"Most of the points I have described here are obvious and easy to find so they should be fully aware from day one. So I'm curious why they have done this kind of work — very unusual for ExxonMobil," he said.
ExxonMobil has referred all questions about the project documents to AGDC since the state-owned corporation took over Alaska LNG.
AGDC pointed to ExxonMobil when it was discovered that maps developed early in the project work outlining prospective LNG plant sites marked an incorrect location for Point MacKenzie, near the port the Matanuska-Susitna Borough has advocated as a potential site for the plant.
Carlstrom said via email that the designs in question were done by a joint venture between contractors CB&I and Chiyoda Corp during the two-year-plus pre-front end engineering and design, or pre-FEED, period that ended when AGDC took over the project in January 2017. He added that the designs are current but they will likely be updated as the last phase of the project before an investment decision is made and when performance guarantees are made.
Economics of gas treatment
Finally, Akiyama said he wonders whether or not the project can be economic in part because of the high content of carbon dioxide in the raw natural gas and the North Slope location of the gas treatment plant, or GTP.
The carbon dioxide, upwards of 10 percent, must be stripped out and sequestered back in the reservoir at the GTP.
Usually, gas treatment and LNG plants are integrated to take advantage of operational efficiencies.
"We need a lot of energy to pull carbon dioxide out. So GTP consumes a lot of energy, also," he said.
Traditionally, GTP plants use waste heat from the gas turbine engines that drive the LNG compressors to help in the process of stripping the carbon dioxide out of the raw gas, according to Akiyama.
Gas for Alaska LNG must be piped as utility-grade gas for use by communities and industrial work along the corridor.
Without the ability to join the plants, the Alaska LNG Project will have to burn comparatively more gas than other LNG plants to power the North Slope GTP, Akiyama said.