America’s New Secret Machine Drill the Mountain Within 10 Days

追加: 2026-06-01

[00:00:00]Beneath the streets of Seattle, between 2013 and 2017, a machine the size of a five-story building cut a tunnel under the city.

[00:00:09]The machine was called Bertha.

[00:00:12]She was a tunnel boring machine, a TBM in industry shorthand, manufactured by the Japanese company Hitachi Zosen for the State Route 99 Tunnel Project that replaced the aging Alaskan Way Viaduct on Seattle's waterfront.

[00:00:29]Bertha measured 17 and 1/2 m in diameter at her cutterhead, approximately 57 and 1/2 ft across.

[00:00:36]At the time of her construction, she was the largest diameter tunnel boring machine ever built.

[00:00:42]She bored through the soft ground beneath downtown Seattle, cutting and removing earth and rock with her rotating face while simultaneously installing the concrete lining of the tunnel behind her.

[00:00:54]Bertha represented the state of the art in American tunneling technology.

[00:00:59]The tunnel boring machine in its modern form is the descendant of 19th-century mechanical tunneling experiments and the more rigorous engineering work of the mid-20th century.

[00:01:10]Particularly, the foundational designs developed by James S. Robbins of the American Robbins Company in the 1950s.

[00:01:19]Tunnel boring machines cut tunnels by mechanical force.

[00:01:24]A rotating cutterhead at the front of the machine fitted with hardened steel cutting tools presses against the rock or soil ahead of it.

[00:01:33]The cutterhead grinds the material away.

[00:01:36]The broken pieces are removed by a conveyor system, and the lining of the tunnel is installed segment by segment as the machine advances.

[00:01:45]The basic concept is unchanged from the 1950s.

[00:01:49]The execution has been refined to exceptional scales.

[00:01:53]Bertha was over 5,000 metric tons of integrated tunneling system, but the underlying method has remained mechanical for seven decades.

[00:02:03]This is the technology that has built most of the great tunnels of the modern era.

[00:02:09]The Gotthard Base Tunnel through the Swiss Alps at 57.1 km is the world's longest currently operational railway tunnel.

[00:02:19]It was completed in 2016 after 17 years of construction, and it was built primarily with tunnel boring machines.

[00:02:26]The Brenner Base Tunnel, currently under construction between Austria and Italy, will be even longer when complete at approximately 64 km.

[00:02:36]The East Side Access project in New York, which brought the Long Island Railroad into Grand Central Terminal, used tunnel boring machines through more than 20 years of construction at a cost exceeding $11 billion.

[00:02:50]Every major tunneling project in the modern industrial world uses some version of the tunnel boring machine.

[00:02:58]But the technology is slow and expensive.

[00:03:02]Standard tunnel boring machine tunneling costs approximately $100 million to $500 million per mile in the United States, depending on geology, diameter, and depth.

[00:03:14]The machines themselves cost tens to hundreds of millions of dollars and require enormous logistical operations to assemble, deploy, and maintain.

[00:03:25]When a tunnel boring machine advances at, say, 15 m per day in difficult ground, the speed of underground construction is limited by the speed of the machine, and the speed of the machine is limited by the physics of grinding rock with mechanical tools.

[00:03:40]There is only so fast a steel cutter head can press against granite before something has to give.

[00:03:45]The mechanical limit is a real engineering constraint. For decades, the tunneling industry has accepted these limits as more or less fixed.

[00:03:53]Various incremental improvements have been made. The machines have become more sophisticated. Cutter head designs have been refined.

[00:04:01]But the fundamental method, pressing steel against rock and grinding the rock away, has remained the same.

[00:04:09]Now, two American startups are trying to change this fundamentally.

[00:04:14]They are not trying to make better TBMS.

[00:04:17]They are trying to replace the entire approach.

[00:04:20]And the technology they are proposing to replace it with is, on its face, almost science fictional.

[00:04:27]Plasma torches.

[00:04:29]What EarthGrid is actually [music] trying to do to replace 70 years of mechanical tunnel boring with plasma torch excavation.

[00:04:36]Why a Korean sovereign wealth fund entity has committed $18 billion to underground projects built on a technology that has not yet been delivered at commercial scale.

[00:04:47]Where the gap between demonstrated field tests and the headline claims leaves the actual state of the work in late 2025.

[00:04:56]The more prominent of the two companies is called EarthGrid PBC, based in Richmond, California.

[00:05:03]The company was founded in 2016.

[00:05:06]Its founder and CEO is a man named Troy Helming.

[00:05:11]Helming had previously founded two renewable energy companies before he turned to tunneling.

[00:05:17]He has been widely quoted in technology and business press as the public face of plasma tunneling.

[00:05:24]The pitch goes like this.

[00:05:26]Mechanical cutter heads have been the dominant tunneling technology for 70 years because no better alternative was available.

[00:05:33]But the physics of breaking rock does not actually require mechanical force.

[00:05:38]Rock breaks under thermal shock as well.

[00:05:41]If rock is heated rapidly enough, the surface layer expands faster than the deeper layer.

[00:05:47]The differential creates fractures that propagate through the rock and cause pieces to break off.

[00:05:53]This is the same physics that causes glass to crack when hot water is poured into a cold cup, scaled up to industrial application. The process is called spallation. And in principle, directing enough heat at a rock face in a controlled way could break rock with heat rather than with steel.

[00:06:12]What plasma tunneling is actually proposing to replace is the part of the story most coverage has never been given.

[00:06:19]Not better tunnel boring machines. The entire 70-year-old approach of grinding rock with steel.

[00:06:26]This is the central technical idea behind EarthGrid's approach.

[00:06:30]The company's rapid burrowing robot, or RBR, uses plasma torches at temperatures reportedly reaching approximately 48,600° Fahrenheit, or around 27,000° Celsius.

[00:06:46]Plasma is the fourth state of matter.

[00:06:49]A gas so hot that its atoms have separated into ions and free electrons.

[00:06:54]Plasma torches direct streams of this superheated gas at a target.

[00:06:59]In EarthGrid's design, the streams are directed at the rock face of a tunnel.

[00:07:04]The thermal shock fractures the rock surface, and the fragments, described in EarthGrid's materials as cornflake-size bits, break away and are removed.

[00:07:14]The company's marketing claims for what this technology can do are bold.

[00:07:19]EarthGrid has stated in industry interviews and company materials that its technology could be 100 times faster and 98% cheaper than conventional tunnel boring machines.

[00:07:30]CEO Troy Helming has cited potential tunneling speeds of 400 to 1,200 m per day in some geologies, compared [music] to the 15 m per day baseline of many tunnel boring machine operations in difficult ground.

[00:07:46]He has described the plasma torches in one widely quoted image as lightsabers vaporizing rock.

[00:07:53]It is important to be honest about the status of these claims. They are company marketing and CEO interview figures.

[00:08:00]They are not independently verified at commercial scale.

[00:08:03]As of late 2025, no full-scale plasma bored tunnel has been delivered to a paying customer.

[00:08:10]The demonstrated technology is at the field test stage, not at the full mile of finished tunnel stage.

[00:08:17]The headline figures, 100 times faster and 98% cheaper, are what EarthGrid believes it can achieve if its technology works at scale.

[00:08:26]The work to demonstrate whether it actually can is still in progress.

[00:08:32]But the company has made real progress, and the support behind it is genuine.

[00:08:37]In June of 2024, EarthGrid announced that its tunnel boring robot had completed successful field tests >> [music] >> using a newly developed two-torch plasma excavation system.

[00:08:49]The tests were described as having taken place in some of the most challenging tunneling geologies.

[00:08:55]The two-torch system represents a step up from earlier single-torch designs, and the field test was a real demonstration.

[00:09:03]Though at a scale far smaller than what would be required for a major infrastructure project.

[00:09:08]The funding history tells part of the same story.

[00:09:11]By June of 2024, EarthGrid had raised more than $30 million in seed funding, doubling its initial $15 million target.

[00:09:20]By mid-2025, the total venture funding had reached approximately $63 million.

[00:09:28]This is real money. Venture capital does not flow to ideas without some confidence that the underlying technology has potential.

[00:09:36]On October 4th, 2024, EarthGrid won the Startup World Cup 2024 grand prize, a $1 million award given at a major international startup competition. This was a significant industry recognition, and in the most consequential commercial development to date, EarthGrid announced a joint venture with Enertech, an entity of the Korea Investment Authority, a Korean sovereign wealth fund organization, for $18 billion in committed United States underground utility projects.

[00:10:09]$18 billion.

[00:10:11]This is one of the largest infrastructure commitments to a tunneling startup in history.

[00:10:17]It represents a bet by a major institutional investor that EarthGrid's technology will be able to handle large-scale underground deployment.

[00:10:25]And it provides the company with the kind of capital backstop that very few startups in any sector ever achieve.

[00:10:33]In parallel, EarthGrid has secured utility approvals in a growing number of states in the United States.

[00:10:40]Originally including New York, Florida, and Washington, the company's regulatory approvals expanded to over 46 states by 2025.

[00:10:51]State utility approvals give EarthGrid streamlined right-of-way access for tunnel deployment.

[00:10:57]This is a significant regulatory advantage that allows the company to move more quickly through the permitting process than a non-utility company could.

[00:11:07]The market the company is positioning toward is real and large.

[00:11:13]The driving motivation, in plain terms, is wildfire risk. California has experienced repeated catastrophic wildfires over the past decade. Many of them ignited by overhead power lines.

[00:11:26]Pacific Gas and Electric, the major California utility, has faced enormous liability and reputational damage from fires caused by its overhead infrastructure.

[00:11:37]The state has been pushing toward burying power lines underground for years, but the cost of doing so with conventional tunneling technology has been prohibitive at the scale required.

[00:11:47]Burying thousands of miles of transmission and distribution lines at conventional costs of 100 million to 500 million dollars per mile runs into the trillions of dollars.

[00:11:59]The math has not worked.

[00:12:01]This is the market EarthGrid is aiming at.

[00:12:04]If the company's plasma approach can deliver tunnels at the cost reduction it claims, even at a fraction of the headline 98% claim, then burying America's power grid underground becomes economically feasible in a way it has never been before.

[00:12:19]EarthGrid is not the only company pursuing this.

[00:12:22]The primary competitor in the United States thermal tunneling space is a California startup called Petra, founded in 2018, which has raised funding from venture capital firms, including Founders Fund.

[00:12:36]The presence of two well-funded American startups pursuing the same fundamental technological approach suggests that the broader investor community sees real potential in thermal tunneling.

[00:12:49]Beyond the plasma startups, there is the broader American tunneling landscape that frames why all of this matters.

[00:12:57]The Boring Company, founded by Elon Musk in 2016 and headquartered in Las Vegas, has been pursuing a different approach to tunneling disruption, not through fundamentally new technology, but through smaller diameter tunnels and faster mechanical TBM design.

[00:13:15]The company operates the Las Vegas Convention Center Loop, which opened in June of 2021.

[00:13:22]The Loop is an underground transit system using Tesla vehicles in two parallel tunnels, each about 8/10 of a mile long, connecting halls of the convention center.

[00:13:33]The Boring Company has been approved [music] to expand this system to approximately 68 miles of tunnels and 104 [music] stations across Las Vegas.

[00:13:43]Though as of 2025, only a fraction has [music] been built.

[00:13:48]The Boring Company's current generation Prufrock tunnel boring machine or TBM advances at approximately a quarter of a kilometer per day with a stated goal of eventually reaching about 1.13 kilometers per day.

[00:14:03]These are substantially faster rates than traditional TBMs in similar geology achieved through smaller tunnel diameter and various mechanical refinements.

[00:14:14]But they remain solidly within the framework of mechanical excavation.

[00:14:18]The Boring Company is making conventional TBM technology faster.

[00:14:23]It is not replacing the underlying approach.

[00:14:26]The plasma startups are different.

[00:14:29]They are not trying to make TBMs better.

[00:14:31]They are trying to make TBMs obsolete.

[00:14:35]If thermal tunneling works at scale, it does not improve on mechanical tunneling. It bypasses it entirely. The cutter head and the steel and the grinding all become unnecessary. The plasma torches do the work the cutter head used to do faster and at less mechanical cost.

[00:14:51]What makes EarthGrid's story particularly compelling, beyond the technology itself, is the personal trajectory of Troy Helming.

[00:14:59]The CEO has been candid in industry interviews about how difficult it was to raise the funding the company eventually obtained.

[00:15:06]Helming pitched investors more than 2,000 times. He has described the rejection rate as approximately 97%.

[00:15:14]Most of the venture capital community did not believe in plasma tunneling. He kept going and eventually the funding came. First, the $30 million in seed.

[00:15:24]Then the additional capital that brought the total to $63 million by 2025.

[00:15:30]Then the Startup World Cup win.

[00:15:32]Then the $18 billion Enertech joint venture.

[00:15:36]The trajectory from being told no 2,000 times to securing one of the largest tunneling commitments in history is, in its own right, a notable American entrepreneurial story.

[00:15:47]Helming has been featured in Wired magazine and other major outlets.

[00:15:52]But the honest assessment of where the technology actually stands is important to maintain.

[00:15:59]The mainstream tunneling industry, represented by organizations like the US National Underground Construction Association and publications like Underground Construction and the American Society of Mechanical Engineers tunneling coverage, has treated plasma tunneling as promising but not yet proven at scale.

[00:16:16]The two-torch field test in June 2024 is a real demonstration, but it is not a full-scale operational deployment. The 400 to 1,200 m per day claim is real as a company claim, but it has not been independently verified in a commercial tunneling project.

[00:16:34]The 98% cost reduction is the company's projection of what scaled deployment would achieve, not a benchmarked result from completed work.

[00:16:43]The 10 days through a mountain framing that sometimes circulates in popular coverage is an extrapolation from the upper end of EarthGrid's claimed speeds.

[00:16:53]If a plasma boring system could sustain 1.2 km per day, it would traverse a 12-km mountain in 10 days.

[00:17:01]The arithmetic works, but the demonstrated technology has not yet been shown to sustain that speed across a full-scale tunneling operation.

[00:17:10]The claim is plausible given the underlying physics and the early results, but it has not been demonstrated.

[00:17:17]No machine has drilled through a mountain in 10 days.

[00:17:21]The capacity to do so is what the company is working toward, not what it has accomplished.

[00:17:28]This distinction matters.

[00:17:31]The gap between what plasma tunneling could do in principle and what it has actually demonstrated in practice >> [music] >> is the central question for whether the disruption EarthGrid is pursuing will actually happen.

[00:17:45]If the company can scale its field test results to the kinds of speeds it has projected, and if it can do so at the cost reductions it has projected, then American underground construction will be fundamentally transformed in the next [music] decade. Power lines could be buried cheaply.

[00:18:03]Fiber optic networks could be deployed underground at scale.

[00:18:08]Water and sewer infrastructure could be installed without surface disruption.

[00:18:13]The Enertech joint venture, worth $18 billion, is a bet that this transformation will happen.

[00:18:20]If the technology does not scale as projected, the bet will fail.

[00:18:25]The cost reductions will not materialize.

[00:18:28]The speed claims will remain claims.

[00:18:31]Plasma tunneling will join the long history of promising tunneling technologies that did not deliver on their early claims.

[00:18:40]What is genuinely happening as of late 2025 is that the question is open.

[00:18:46]EarthGrid has the funding to develop the technology further. It has the regulatory approvals to deploy it across most of the United States. It has a major joint venture providing committed work and capital.

[00:19:00]The next several years will determine whether the technology actually scales.

[00:19:06]In the meantime, the broader context of American tunneling continues.

[00:19:11]The Boring Company continues operating the Las Vegas Convention Center Loop and pursuing its expansion plans.

[00:19:19]Conventional tunnel boring machine operations continue on major projects across the country and the world using machines from manufacturers like the Robbins Company in Ohio and the German firm Herrenknecht AG.

[00:19:33]Herrenknecht is the world's largest tunnel boring machine manufacturer and supplied many of the largest American tunneling projects including Bertha and the East Side Access Project.

[00:19:46]The fundamental technology of the modern tunneling industry, mechanical tunnel boring machines grinding through rock with steel cutter heads, continues to be the workhorse of underground construction worldwide.

[00:19:59]The plasma startups represent a potential alternative future. They are not the present.

[00:20:05]The present remains mechanical, expensive, and slow by the standards EarthGrid is trying to redefine.

[00:20:12]But the potential future, if the plasma approach scales, would be genuinely different.

[00:20:19]Tunneling that costs a fraction of what it costs today.

[00:20:23]Tunneling that proceeds at speeds that current methods cannot match.

[00:20:28]Underground construction available for projects that have historically been priced out of feasibility.

[00:20:35]Power lines buried at scale to address wildfire risk.

[00:20:39]The wildfire driver is the largest single market force behind the entire effort.

[00:20:45]California's repeated catastrophic wildfires have caused immense damage, billions of dollars in losses, dozens of deaths, and the destruction of entire communities.

[00:20:56]A significant fraction of these fires have been ignited by overhead power lines, particularly during high wind conditions when sparking from damaged equipment can start fires that spread rapidly through dry vegetation.

[00:21:10]Pacific Gas and Electric filed for bankruptcy in 2019, in significant part because of wildfire related liabilities.

[00:21:19]The straightforward solution, burying the power lines underground, has been technically possible all along.

[00:21:27]What has been the obstacle is cost.

[00:21:30]Burying power lines at scale, using conventional tunneling and trenching technology, is prohibitively expensive across the thousands of miles of high fire risk corridors that California has.

[00:21:42]The state has been pursuing burial in selected high priority areas, but a comprehensive solution at the scale of the actual problem has not been economically feasible.

[00:21:53]This is the gap that plasma tunneling, if it works as projected, would close.

[00:21:58]If EarthGrid or Petra can deliver underground utility tunnels at a small fraction of conventional cost, the math for burying the California power grid changes fundamentally.

[00:22:11]The state could afford to do at scale what it can currently only afford to do in selected corridors.

[00:22:18]This is what the $18 billion EnerTech joint venture is, in effect, a bet on.

[00:22:24]The committed dollars represent confidence that the technology will scale, and that the market will be there when it does.

[00:22:31]Whether that confidence proves justified will depend on the engineering work of the next several years.

[00:22:38]For now, what we have is a real company with real funding pursuing a real technological approach to a real market opportunity.

[00:22:47]The plasma torches reach the temperatures the company has claimed.

[00:22:51]The spallation physics works as described.

[00:22:55]The two torch field test was conducted as announced.

[00:22:58]The funding and the joint venture and the regulatory approvals are all real.

[00:23:03]The marketing claims about speeds and costs are not yet verified at commercial scale, but they are also not implausible given the underlying engineering.

[00:23:13]What is not real is any specific demonstrated case of a machine drilling through a mountain in 10 days.

[00:23:21]That is the upper end of company projections, not a delivered result.

[00:23:26]The actual story in honest framing is more interesting than the simplified version anyway.

[00:23:32]A California startup founded in 2016.

[00:23:35]A CEO who pitched investors more than 2,000 times with a 97% rejection rate.

[00:23:42]$63 million in venture funding raised.

[00:23:46]The Startup World Cup 2024 Grand Prize won. An $18 billion joint venture secured with a Korean sovereign wealth fund entity.

[00:23:57]Utility approvals obtained in 46 states across the United States.

[00:24:03]Plasma torch tunneling technology now in development could, if it scales as projected, replace 70 years of conventional mechanical tunnel boring with a fundamentally different approach to underground excavation.

[00:24:16]That is what EarthGrid is. That is what it is trying to do.

[00:24:20]And the $18 billion bet from EnerTech, combined with the institutional support from the broader venture and infrastructure communities, suggests that serious people with serious money believe the bet is worth making.

[00:24:34]The tunnel boring machine, in its conventional form, will continue to be the workhorse of global underground construction for the foreseeable future.

[00:24:43]Bertha is still under Seattle, having completed her work.

[00:24:47]The Brenner Base Tunnel will be completed by conventional tunnel boring machines.

[00:24:52]The next great Alpine railway tunnels will likely use the same fundamental technology that built the Gotthard.

[00:24:59]The 70-year history of mechanical tunneling will continue for some time yet.

[00:25:04]But somewhere in California, in Richmond, in EarthGrid's facilities, plasma torches are being refined. The two torch system is being improved. The next generation of field tests is being prepared.

[00:25:18]The $18 billion Enertech commitment is providing the funding to push from field test scale toward operational deployment.

[00:25:26]The company believes, and its investors apparently believe with it, that within the next several years, plasma tunneling will move from being a promising research and development effort to being a deployed commercial technology that begins the actual displacement of mechanical tunnel boring machines from major American underground construction.

[00:25:47]If that happens, the resulting transformation of American infrastructure will be significant.

[00:25:53]If it does not, plasma tunneling will join the longer list of promising tunneling innovations that did not scale.

[00:26:00]The next several years will tell us which outcome is the right one. That is the actual story, >> [music] >> not a secret machine that has already drilled a mountain in 10 days.

[00:26:10]A real, well-funded, increasingly serious American attempt to replace 70 years of mechanical tunneling with a fundamentally different technology.

[00:26:20]Backed by venture capital, a Korean sovereign wealth fund joint venture, and a regulatory framework that gives the lead startup utility access across most of the United States.

[00:26:32]The plasma torches are real. The funding is real.

[00:26:35]>> [music] >> The market is real.

[00:26:37]The engineering challenge of scaling from field tests to operational deployment is also real, and it remains the central open question.

[00:26:45]What EarthGrid is attempting is the most significant attempted disruption of tunneling technology since James S.

[00:26:51]Robbins built the first modern tunnel boring machine in the 1950s. [music] Whether it succeeds will determine whether the underground geometry of American infrastructure in the 2030s looks fundamentally different from the underground geometry of the present.

[00:27:07]For now, what is true is what is documented.

[00:27:11]A California startup is trying to replace tunnel boring machines with plasma torches.

[00:27:16]It has the funding, the partnerships, the regulatory approvals, and the institutional support to attempt the transition.

[00:27:24]The technology has been demonstrated at field test scale >> [music] >> and is being prepared for commercial deployment. The market, driven by wildfire risk, by aging infrastructure, by the broader pressure to bury American utilities underground, is large enough to support a successful disruption if the technology delivers.

[00:27:44]What is not yet true is that any machine, plasma or otherwise, has drilled a mountain in 10 days.

[00:27:51]The capacity for that kind of performance is what the company is working toward, supported by genuine engineering effort and serious institutional backing.

[00:28:01]The capacity has not yet been demonstrated.

[00:28:04]That is what the genuine American tunneling story looks like in late 2025.

[00:28:10]A revolutionary effort underway.

[00:28:12]A historic commitment of capital.

[00:28:15]A real but still developmental technology. A market poised for transformation.

[00:28:21]And a CEO who pitched investors 2,000 times before getting the funding to try to underground America's power grid with what he has described as 10 lightsabers vaporizing rock.

[00:28:35]The plasma torches are real.

[00:28:37]The mountain has not yet been drilled in 10 days.

[00:28:41]The next chapter is being written.