It begs the question - why use solar if the storage mechanic is sufficient to cover peak/off-peak storage of any surplus electricity in a significant way? Hydro Quebec has significant ($500M+/yr) electricity surpluses on a given year in the form of underutilized hydro during spring/fall months where energy demand is very low and export possibility is limited. These seem like ideal candidates for this vs something like a (relatively) small-scale solar setup. The slow-charge slow-release is also a considerable factor here since most battery energy storage solutions are designed for charge/discharge events multiple times per day to arbitrage energy rates - their ROI (which is currently very aggressive, sometimes less than 4-5 years) is a function of how many charge-cheap/discharge-expensive cycles they can perform per year. That is something that this doesn't seem equipped to do which is a negative multiplier on the ROI these systems can yield. If you can charge/discharge a battery 500 times per year, and the cost is 500x more, this is only at parity with a much slower charge/discharge schedule (charge during the summer, discharge during the winter - or vice versa). The other reality here is the steam turbine required to generate electricity (assuming electricity storage and not heat storage model) is a significant capex - roughly $300-400/kW (and some considerable OPEX in the form of maintenance and regulation/compliance). In this reality it only makes sense to retrofit an existing sunken-cost turbine and grid transmission system vs purchasing and co-locating a generating turbine along with the storage solution.

For reference, point-in-time energy market rates usually swing by 2x-3x per day - meaning if you charged during the cheapest market rate and discharged during peak you'd net about 2.5x return on that cycle) - even more so during extreme temperature events like heat waves or cold freezes - those are ultimately what you're riding here in terms of validating the system's viability from a financial perspective. If you reduce that scale from hours to months, and if draw-down speed is slow (ie: you can't sustain 50MW of steam with 500,000 tons of dirt even at 600'C) then you're looking at even more complicated returns.

By my simple, assumption-laden math, a 50MW "system" (capable of providing up to 50MWe peak output and requiring a requisite (assuming since it's not mentioned in the article - that at 200'C a 1,000,000 ton dirt pile would only be able to sustain 40MW of thermal output/20MW of electrical output and 240MW thermal/120MW electrical output at 600'C) would be:

PV system (20MW system would require ~30 days of charging to provide 50MWe output for 1 day, ~1200MWhe), alternatively, per day, you could discharge 50MWe for ~48 minutes. 1,000,000tons of dirt storage at 600C should hold a theoretical ~28 days of 50MW electrical supply. (also worth noting, getting the dirt pile heat up to "steam" temp would likely eat up a considerable number of months charging, which is also capex)

$1,000,000 for dirt

$5,000,000 for balance of system (heater elements and wiring + ASME tubing - as an aside this seems very opportunistic for 20MW of heaters and tubing to supply 100MWt of steam)

$12,000,000 for Solar Panels ($0.60/w bulk)

$8,000,000 for Solar Supporting systems and installation (assuming heaters can run on DC power and no inverters are required and there is no grid tie, minimal permitting and simplified ground install)

$25,000,000 for a 50MW steam generator turbine and transformer yard, provisioning etc

land use: ~25 acres for dirt pile, ~100 acres for solar, 10 acres for steam/aux, call it $300,000 assuming US averages for cleared land.

----------

Assuming north-eastern US (~20% solar efficiency with subzero winters where you also have high off-solar peak demands)

If you only charge/discharge this twice per year you're looking at some pretty paltry economics - you could only really fill about 18k MWh of thermal energy during half of the year for a ~7,400MWhe discharge - $592,000 gross electricity revenue per discharge cycle at an opportunistic 7-day "peak" market rate of $80/mWh which is about $1,184,000/yr gross margin. If you did it once per day (40MWhe per day at peak average intra-day market rate -$68/mWh) you're looking at ~$2,720/day or $992,800/yr gross margin.

$51M capex would be difficult to justify margins of only $1.1M/yr, and that's before any operating costs of which there would be several.

If you just sold the same solar at market rate (~$36/MWhe) throughout the year you'd net out at $1,261,440. Capex would be ~$40M and grid-tie solar is very cost effective in OPEX.

Likewise, if you just connected the system to the grid and skipped solar altogether (powering the heaters with grid energy like battery storage would): 50MW in for 12 hours on cheap time-of-day rates (typically overnight ~$18-20/MWh) and sold for 5 hours during daily peak rates ($55) you'd cut your capex considerably without the solar component and you'd be able to net, even with round-trip energy efficiency around 41%, (600MWe in @ $11,400, 248MWe out @ $13,640 = $2,240/day ~$817,600/yr gross margin) for a capex of $31.3M.

So in the end, the best solution seems to be collocating this on an existing coal/gas plant, where the capex is already sunk in the transformers, grid interconnect, steam turbine, land and permits and you're only adding the earth battery - you could run the model with the above margins with a capex of only $6-7M, which is very viable and even more favorable than the economics of spinning up a new gas/coal plant.

The economics of battery energy storage (BES) systems are much better known (ROIs of <4 years in extreme-swing energy markets doing intra-day peak arbitrage is very possible) since your round-trip efficiency is closer to 91%. A 250MWh BES plant with 1-hour charge/discharge window would be~ $40M installed and could arb twice per day - at 2x (low end averages - buying at $26 and selling at $52 twice per day = $14,285 cost for $26,000 revenue) $11,715 margin per day, $4,275,975/yr on $41M capex is still better economics than all the above models except those where the steam generator and grid infrastructure is already sunk.

If you had more centralized heating systems (like https://en.wikipedia.org/wiki/Seattle_Steam_Company or https://en.wikipedia.org/wiki/Drake_Landing_Solar_Community ) then put the extra heat in there for use in winter.

Imagine 1,000,000 Drake Landing installations per year in Canada, pre-heating with the excess electricity. In 30 years Canada would need zero fossil fuels for buildings.

which... is only 13% of their GHG emissions? Oh we're fucked. The planet's so fucked.