Description

Perma-Fix (PESI) operates 4 of the 10 active mixed level nuclear waste treatment facilities in the US. These are scarce assets that face NIMBY opposition, making them difficult to replicate. For context, the last facility was built 20 yrs ago. Permits are administered at the federal, state, & local level & cover everything from environment remediation to worker safety. Obtaining all the NRC, EPA, state, & local permits needed to build a plant takes 5-7 yrs w another 3-4 yrs needed to finish construction. In addition, the plant has to be in close geographic proximity to existing radioactive waste storage sites to minimize the risk of spills. Government departments are not interested in the financial & legal liability exposure associated w transporting radioactive waste streams long distances via truck or rail. PESI estimates the replacement cost of their fully permitted facilities is $200M compared to a $166M enterprise value.

Hundreds of different waste streams w diverse radioactive content profiles each require individual permit approvals that take bw 6-24 months to secure. The unique radioactive characteristics of these waste streams also requires the development of highly customized treatment processes using domain expertise built through yrs of prior experience. This doesn't favor new entrants. PESI has over 50 different patent-protected chemical processes tailored to the classification, mixing, & processing of individual waste streams from a wide range of storage sites & has leveraged decades of technical knowledge to build competency in processing the most difficult types of waste. For example, they operate the only mixed waste thermal combustion unit in the US at one of their plants. Doing so requires constructing a detailed understanding of the composition & intensity of various radioactive compounds within each waste stream as well as filtering out certain elements, processes they've spent decades perfecting. Stringent regulatory guidelines, limited domain expertise, & lengthy waste-specific permit lead times naturally constrain the volume of waste a new plant can initially treat. This means the 10-yr construction lead time is typically met w a small initial rev stream paired w a sizable fixed cost structure. I estimate it would take a new entrant >15 yrs to breakeven, explaining why no one has even tried to enter this market in the last 2 decades.

These factors create stringent entry barriers that preserve the throughput volumes PESI treats at their existing plants. Their established reputation & impressive track record also positions them to treat the large backlog of radioactive waste stored in several decommissioned nuclear sites. Hanford is the largest such opportunity, w 56M gallons of radioactive waste sitting in underground single & double shell metal tanks 40+ yrs past their useful design life. Hanford represents a growing $280B environmental liability for the DOE. Waste classification informs the treatment process & is predicated on the intensity & composition of the various radiological compounds stored in the tanks. 50M gallons are classified as low activity waste (LAW), w the remaining 6M being high level waste (HLW). The tanks have been underground for 60+ yrs, which has caused the waste to solidify into a sludge. The DOE must retrieve this sludge from the tanks by splicing it w water & chemicals, pretreating it (ie separating the low radioactivity compounds from the high ones), then immobilizing it.

The DOE is obligated to start treating Hanford LAW by August 2025 under the Tri-Party Agreement, a contract they signed w Washington state. After 20 yrs of cost overruns & delays, construction of the direct feed low-activity waste (DFLAW) plant was completed in early 2023. This facility will vitrify ~25M gallons of LAW over its useful life but can only address half of the DOE’s LAW treatment liability over the course of the project timeline (>50 yrs). Vitrification is the process of mixing the waste w glass forming materials at 2100º F in 2 big melters. The glass immobilizes the radiological compounds, allowing for safe underground storage w no risk of groundwater/soil contamination. The DOE has spent $14B over the last 12 yrs to build the DFLAW & effluent retention plants & will spend another $4.5B p/yr in ongoing operating costs to run these facilities at maximum allowable capacity. The plant will start cycling through batch volumes in late 2024 & be fully operational (“hot commissioning”) by 2025.

The facility’s annual throughput volume is 1M gallons of LAW p/yr, implying an avg treatment cost p/gallon of $4.5K. Depreciating the upfront construction costs of the facility over aggregate treatment capacity adds another $560 of costs p/gallon ($14B/25M cumulative gallons). The GAO states DFLAW treatment costs p/gallon approximate ~$2.5K, w this estimate excluding waste retrieval, pretreatment, & depreciated plant capital costs. Tank waste needs to be retrieved from tanks & pretreated prior to being vitrified in the DFLAW plant. Pretreatment involves separating the sludge within the tank into low & high activity liquid waste streams. It’s a 4-step process:

1. Retrieve dried waste (“sludge”) from the tanks by mixing it w water & chemicals to create a more soluble “slurry” or “supernatant”

2. Separate the liquid slurry, which contains less radioactive compounds (LAW) from the remaining high activity solids still stuck to the tank

3. Pretreat LAW slurry using a tank side cesium removal system (TSCR) which removes cesium 137 from the liquid slurry to make it suitable for vitrification in the DFLAW plant. Cesium-laced water can’t be vitrified bc it emits highly toxic vapors when exposed to high temperatures that are difficult to capture & condense

4. Treat remaining high activity solid sludge in tank by adding more water/chemicals during “washing/leaching” process

Safely retrieving a gallon of radioactive tank-waste requires adding at least 2-3 gallons of clean water (some experts believe it could be bw 6-10 gallons p/retrieved gallon depending on the tank contents). This generates additional LAW water that needs to be safely immobilized. The DOE classifies this as LAW byproduct or “effluent.” 2 more gallons of effluent LAW are also generated for each vitrified gallon of tank-based LAW. Toxic gases emitted by the melter are captured in an offgas treatment system that condenses the vapor into a liquid that needs to be safely disposed of. So treating a gallon of tank waste generates 4-5 gallons of incremental LAW in the form of effluent, creating a dilemma for the DOE- the more waste they treat, the larger their backlog of untreated LAW gets.

The DFLAW plant will produce 1.8M gallons of offgas effluent LAW byproduct p/yr when it starts operating at peak capacity in 2025. The pretreatment/retrieval process generates another 2-3M gallons of effluent LAW as well assuming DFLAW throughput averages 1M gallons p/yr. Vitrifying the effluent is not a viable option since it’s 1) too expensive & 2) would produce 2 more gallons of byproduct LAW (due to its excessive water content). The waste particles would just evaporate, be recaptured by the offgas condensation system, & be turned back into LAW water. In effect, the DOE’s treatment liability grows as they vitrify more waste. Luckily, they’ve found a cheap, effective solution- shipping the effluent LAW by rail to PESI’s Richland facility 11 miles away where it will be grouted, a process that sets the LAW into a concrete mixture, applies chemicals to it, & then lets the concrete harden. Richland is the only viable option for offsite grouting. The next closest treatment facility w the requisite permit approvals is 1600 miles away in Clive, TX.

The unprocessed waste would have to be trucked to the TX site, a more dangerous transportation method. The DOE estimates it would take 10 yrs to secure the state & local permits needed along the interstate trucking route to TX, making this option not feasible from a legal, financial, & logistical perspective. The proximity of the Richland facility to Hanford coupled w its direct rail connection has provided PESI w monopolized access to a large, highly visible waste stream. Grouting immobilizes the waste by converting it from an aqueous liquid slurry/supernatant into a solid at a much lower cost compared to vitrification. PESI will charge just $45 p/gallon to treat the effluent waste stream compared to the GAO’s budgeted vitrification cost of $2.5K p/gallon. The DOE signed a record of decision to ship 1.8M gallons of effluent p/yr to Richland for the next 10 yrs starting in 2025. The length is a formality- PESI will treat all the effluent produced over the life of the DFLAW plant since there is no alternative way to treat this waste. GAO projects the cost savings generated by grouting just the offgas effluent waste stream amount to >$20B over the 50+ yr timeline.

PESI will generate $80M of annual rev treating the effluent LAW byproduct generated during the vit process, compared to their consolidated $70M of total rev this past yr. The DFLAW plant will run for a minimum of 25 yrs, providing PESI w a highly visible rev stream over that time. Contribution margins avg 70%, w $3M of additional variable labor costs needed for every $10M of added treatment rev. PESI plans to add ~$10M of SG&A overhead as well to support the higher treatment volumes, resulting in $46M of cash EBIT. Modest p/gallon price increases of 5-6% p/yr translate into 7-8% EBIT growth due to fixed cost leverage. The advantaged geographical location of PESI’s Richland plant & the price umbrella resulting from vitrification’s much higher p/gallon treatment costs should offer modest pricing power. For context, even if PESI doubled prices p/gallon to $80 over 10 yrs (6% CAGR), grouting at their plant would still be 30x less expensive than vitrification.

The DOE also plans to ship tank-based LAW to Richland for offsite grouting. This “supplemental LAW” is the 25M gallons of excess waste not treatable by the DFLAW plant under the approved timeline. Retrieving each gallon of tank waste also results in at least 2 gallons of additional LAW water, which I referred to as “pretreatment” effluent byproduct before. That pins the total treatment obligation at 75M gallons. PESI will likely grout the byproduct waste as well. Of note, several technical documents cite the need for 6-7 gallons of supernatant chemical water to retrieve 1 gallon of tank-based LAW. This number will vary on a tank by tank basis & is dependent on the composition of radiological content & other factors. I mention this to 1) highlight the potential upside in total treatable LAW that PESI has exposure to & 2) frame how conservative my FCF assumptions are. The need to immediately treat 25M gallons of supplemental tank-based LAW through an alternative, off-site treatment method doesn’t only arise from grouting’s technical simplicity & much lower p/gallon treatment costs. It’s also critical to building a backlog of HLW to feed into the HLW vit plant.

Remember, low activity & high activity radioactive compounds are tightly mixed as a solid sludge inside the tanks. The pretreatment process described above separates the less radioactive compounds (cesium 137 & strontium 90) from the more dangerous compounds w higher toxicity levels & longer half lives (plutonium 239 & iodine 129) by adding a mix of water & chemicals (called supernatant) to the sludge to create a liquid slurry. The retrieved liquid waste is passed through the TSCR facility & then fed into the DFLAW melters where it’s mixed w glass forming materials at very high temperatures. The remaining tank solids are high level wastes that require more sophisticated treatment/handling, which involves adding more water & chemicals in a leaching process.

The original plan was to feed the combined tank waste into a pretreatment facility that could separate it into low & high activity mixes by measuring radiological content & particle size while simultaneously splicing it w supernatant. The DOE ran into so many technical problems trying to develop this facility that they stopped designing it in 2012 & have no plans to resume. A major risk involved larger plutonium particles getting attached to the pulse jet mixer when it was actively separating the low & high activity compounds, which would catalyze a nuclear chain reaction.

The DOE is legally obligated to start treating the HLW by 2035. Running the HLW vit plant at full capacity requires having enough separated HLW to feed into the melters by 2033 (when the plant will be finished). Building an adequate HLW backlog entails retrieving & treating millions of gallons of LAW first, bc it’s easier to pull this waste from the tanks (it comes out first) & there isn’t enough tank storage space to build an untreated LAW backlog. Failing to separate the LAW compounds from the mixed sludge in a timely manner would jeopardize the HLW treatment timeline & leave highly radioactive materials in deteriorating underground tanks w a growing risk of leakage. This dynamic heightens the need for an alternative treatment method for supplemental LAW, providing PESI w an attractive opportunity to increase their Hanford treatment volumes by ~55% (from 1.8M gallons of effluent to 2.8M gallons of combined throughput).

PESI expects to start Phase 2 of the TBI by October 2024, which involves grouting 2K gallons of tank-based supplemental LAW. This proof of concept trial positions them to start treating 1M gallons of supplemental LAW by 2026, w the ability to 3x supplemental LAW treatment capacity w a capex investment of $8M. PESI charges $100 p/gallon for this waste stream rather than $45 bc they perform more complicated leaching & cleaning processes since the waste carries a greater concentration of radioactive elements. They also store it at a 3rd party site.

Shipping & landfill costs eat into the incremental margins of this rev, although they still approach 65%. SG&A overhead is budgeted at $13M, translating into cash EBIT of ~$52M. “Pretreatment” LAW byproduct generated by mixing clean water w tank waste during Step 1 of the retrieval process will likely also be grouted under the “supplemental LAW” contract structure since there isn’t enough landfill or tank space at Hanford to store it & vitrifying it would 1) require building a new DFLAW plant which is financially unfeasible & a massive technical headache 2) result in the generation of 2 gallons of offgas effluent p/vitrified gallon & 3) increase the treatment budget by a factor of at least 25x p/gallon given the disparity in treatment costs bw the 2 waste processing methods. Grouting 2M more gallons of pretreatment LAW byproduct would add at least $100M of incremental tank-based LAW EBIT. I believe PESI has a good chance to capture that volume as well as the byproduct volume generated from retrieving the tank LAW being fed into the DFLAW plant for vitrification (at least another 2M gallons p/yr). In effect, the company could be grouting 6.8M gallons p/yr at some point in the future.

I'm confident by 2027, PESI will conservatively be treating 2.8M gallons of LAW at a blended rate of $64 p/gallon. Hanford rev of $180M translates into $121M of gross profit (67% blended incremental contribution margins) burdened by $23M in SG&A expense. Cash EBIT generated from treating effluent LAW byproduct & supplemental tank-based LAW from the Hanford site is $98M w significant upside based on 1) the opportunity for incremental treatment volumes from retrieval/pretreatment waste streams that I previously covered (at least 2M gallons, maybe 4-5M p/yr) & 2) PESI raising p/gallon prices.

The remaining business runs at breakeven profitability on $70M of rev. So Hanford EBIT flows directly to the consolidated bottom line, as does any added rev from higher throughput or price increases. Assuming flat rev in the core non-Hanford business, the combined company will do $250M of annuity like rev w $98M of EBIT. After-tax profit of $74M directly translates into FCF, since depreciation approximates maint capex & the business has limited working capital requirements. 2027 cash EPS comes in at $5 compared to a $9 stock today. The predictability of PESI’s rev is underpinned by the 50+ yr clean up timeline of the Hanford site & the massive amount of effluent offgas & pretreatment LAW byproduct that will be generated in the process. While there’s a chance PESI doesn’t grow the top line much after the Hanford volumes fully ramp (assuming they can’t take price &/or only process 2.8M gallons p/yr), the company will still benefit from highly visible cash flows that will be used to retire shares. Cash EPS growth will approximate the rolling avg FCF yield assuming all cash goes to repurchases.

If PESI eventually treats 2M additional gallons of pretreatment LAW p/yr, companywide FCF doubles to $150M assuming the same incremental EBIT margin profile of 52%. I expect 1M gallons of added throughput to flow through at the 70% contribution margin, since G&A is very scalable. Still, you don't need that to happen to generate an attractive return. If the company devotes 80% of FCF to repurchases at a blended avg market multiple of 19x, the share count would be >50% lower than today on a similar FCF base of $75M (assumes flat volumes of 2.8M gallons p/yr, split bw 1.8M of effluent at $45 p/gallon & 1M of supplemental tank LAW at $100 p/gallon).

The annual return in that case over 16 yrs comes out to 22% p/yr, w a 2040 stock price of $200. I also assume p/gallon pricing remains flat in perpetuity, an unlikely outcome considering the company's monopolized access to this ever-growing waste stream, the mission-critical & patent-protected nature of the service they provide, & the low cost of the grouting treatment process relative to vitrification. The biggest risk is the DFLAW start time gets delayed, which would push out the effluent LAW rev stream. But this outcome would be offset by the DOE shipping a greater quantity of LAW to the Richland plant for grouting, since the agency is legally obligated to begin treating Hanford tank waste by August 2025 & they have a 900K gallon backlog of pretreated waste in the TSCR facility.

The best case outcome would be for PESI to trade at a very low FCF multiple after the Hanford treatment volumes have fully ramped. This would give management the opportunity to repurchase a lot of shares & supercharge end-state cash EPS levels. There will be a point in time where the market capitalizes PESI’s predictable, high margin, annuity like FCF stream at >30x FCF. The later that is, the higher cash EPS will be, which would translate into a better return. Over a 3-yr timeline, I expect the stock to trade bw $50-100 by 2027 when they earn $5 in EPS. In 10 yrs, PESI could be doing $15 in 2034 FCF/share assuming the share count is down from 16M to 10M & FCF stays flat. A clear margin of safety is demonstrated when considering the replacement value of their plants equates to ~$13 of equity value vs a $10 stock price today.

Catalyst