Every waste infrastructure deal is a bet on a set of assumptions holding: that the tipping fee stays where the pro forma put it, that the tons keep showing up, that a regulator does not rewrite the economics after you close. Scenario modeling is the discipline of taking those assumptions apart before you commit capital, and asking, systematically, what the investment looks like when they move.
Most scenario-modeling guidance is written for generic finance. It teaches you the mechanics of a data table or a probability distribution using a manufacturing plant or a real-estate development as the worked example. That guidance is not wrong, but it is not useful for a landfill, a transfer station, a MRF, or a biogas facility, because it never names the variables that actually break waste deals. Waste is not generic. Three specific sources of uncertainty dominate the return, and a scenario model that does not center them is modeling the wrong thing precisely.
This is the overview article. It maps the three variables that matter, explains the three techniques you have to model them, and tells you which technique to reach for when. Each technique gets a short treatment here and a link to a deeper article. The through-line across all of it: the technique is the easy part, and the inputs are what make or break the analysis.
The Three Variables That Actually Break Waste Deals
Generic scenario modeling asks you to vary “revenue,” “cost,” and “growth.” That framing is too coarse to be actionable in waste. When a waste deal disappoints, the postmortem almost always traces back to one of three variables, and often to their interaction. Before choosing a technique, get clear on which of these your deal is most exposed to.
Tipping-Fee Volatility
For landfills and transfer stations, the tipping fee is the largest line in the model and the least stable. It moves with regional disposal capacity, competitive entry, and regulation. A single competitor opening nearby airspace can compress fees across a market; a capacity constraint can send them the other way. Underwriting a fixed fee for the life of the asset assumes away the single biggest source of revenue risk in the deal.
The mistake is not using a number. The mistake is using a point number when the market has demonstrably produced a range. A credible scenario model treats the fee as something that varies within bounds the specific market has actually generated, not a round-number haircut applied to a national average. Grounding that range in real, facility-level tipping-fee dynamics for the exact market is what separates a scenario that means something from a spreadsheet exercise.
Regulatory Change
Waste is one of the most heavily regulated asset classes in infrastructure, and regulation reshapes facility economics directly. Organics landfill bans divert a slice of inbound tonnage away from disposal. Statewide diversion mandates redirect material flows over a decade. Per-ton environmental surcharges raise the cost side. Extended producer responsibility laws move who pays and how much. Each of these changes the tonnage, the fee, or the cost basis the model rests on.
Regulatory risk is distinctive because it is often the most foreseeable of the three and the most neglected. Bans and mandates are debated, drafted, and phased in over years. The scenario question is directional and structural rather than statistical: if a mandate diverts organics out of your inbound stream, or a surcharge raises your cost per ton, does the deal still clear its hurdle? Because regulation arrives as a discrete event with a probability and a timeline, it is the variable best suited to decision-tree treatment, which we come to below.
Feedstock and Volume Supply Risk
For biogas, composting, and biomass projects, feedstock is the mirror image of tipping-fee risk. On the disposal side, the worry is what you get paid per ton. On the processing side, the worry is whether the tons arrive at all. A dairy digester assumes a herd size and a manure yield. A food-waste facility assumes committed generator contracts hold. A biomass plant assumes residues stay within haul distance and a competitor does not bid them away.
Every one of those is uncertain, and a facility underwritten to its maximum feedstock volume every single year is fragile by construction. Herd sizes change, contracts churn, competing facilities enter the same catchment. Volume risk also applies to throughput on the disposal side: a MRF or transfer station rated for a given tonnage rarely runs at nameplate once you account for downtime, contamination, and seasonal swings. The scenario model has to let volume vary the way the real supply does.
The reason these three deserve to be named individually, rather than folded into “revenue risk,” is that they behave differently and often move together. In a soft market, low fees and light volume tend to arrive at the same time. A scenario model that treats them as independent understates the downside, because the bad cases are correlated. Keep that in mind through everything that follows.
Three Techniques, Three Jobs
There is no single “scenario model.” There are three techniques, and the reason to know all three is that they answer different questions. Using the wrong one either wastes effort or produces false confidence. Here is how each works and when it earns its place.
What-If Analysis: The Fast First Pass
What-if analysis is the simplest of the three and, for most deals, the right place to start. You change one input at a time, hold everything else constant, and observe how the return moves. If dropping the tipping fee by ten dollars turns an acceptable IRR into an unacceptable one, that fee is a load-bearing assumption and deserves serious diligence. If the same swing barely moves the return, you can stop worrying about it.
The most useful structure is the best-case, base-case, worst-case spread: for each critical variable, define a plausible upside, the case you are underwriting to, and a defensible downside, then run the model under each. A deal that only works in the best case is a bet, not an investment. A deal that survives the worst case on every critical variable is one you can size and price with confidence. What-if analysis is cheap enough to run on a facility acquisition in an afternoon, and it tells you which of the three variables above is actually load-bearing for this deal. The full method, including the specific waste scenarios and how to turn findings into decisions, is covered in the what-if analysis deep-dive.
Its limit is built into its strength. By changing one variable at a time, what-if analysis cannot capture what happens when several move together, which, as noted, is exactly how waste deals go wrong. That is the boundary where you graduate to the next technique.
Monte Carlo Simulation: The Combined-Risk Picture
When several variables are uncertain at once and interact, one-at-a-time testing understates the risk. Monte Carlo simulation solves this by varying all the uncertain inputs simultaneously. Instead of a single value, you assign each variable a probability distribution, a range of possible values and how likely each is, and the model runs thousands of iterations, drawing a random value from every distribution each time and recording the result. The output is not a number but a curve: the full distribution of returns the deal could produce, and, critically, the probability it falls below your hurdle.
That probability of loss is the number a base-case pro forma can never give you. But it only holds if the correlation between the variables is modeled deliberately: drawing each variable independently from its own distribution assumes they move independently, which understates the downside. To capture the clustering of low-fee, low-volume scenarios that make the real risk worse than any single variable suggests, the simulation has to impose the relationship explicitly — a correlation matrix, a copula, or a joint distribution that ties the draws together. Skip that step and the model will quietly assume the bad cases never coincide. Monte Carlo is the rigorous second pass, appropriate once what-if analysis has shown that the interactions matter. The mechanics, the choice of distributions, and worked examples for tipping-fee, feedstock, and throughput risk are in the Monte Carlo simulation deep-dive.
Decision Trees: Discrete Events and Sequenced Choices
Some of the most important waste-deal uncertainties are not continuous ranges but discrete events with probabilities and consequences. A permit is granted or denied. An organics ban passes or fails in the next legislative session. An expansion airspace application clears or does not. A decision tree is the tool for these: it maps the branching sequence of events and choices, assigns a probability and a payoff to each branch, and rolls the outcomes back to an expected value for the decision at the root.
Decision trees are especially well suited to regulatory change and to phased, optional investments. If you can defer a second-phase capital outlay until after a mandate is resolved, that option has value, and a decision tree is how you quantify it. Rather than force a “what if regulation changes” question into a single downside case, a tree lets you model the ban passing with some probability, the surcharge landing at one of several levels, and your own response to each, all in one structure. It answers a question the other two techniques do not: given that I can react to how events unfold, what is this decision worth today? For deals where the value hinges on a permit, a vote, or a sequenced build, the decision tree is the technique that fits the shape of the problem.
Choosing Between Them
The three are complementary, not competing. What-if analysis is the fast triage that identifies which variables are load-bearing. Monte Carlo is the rigorous quantification of combined, continuous risk once you know the interactions matter. Decision trees handle the discrete, sequenced, and optional events, regulatory outcomes above all, that the other two model poorly. A disciplined underwriting process often uses all three: what-if to focus attention, Monte Carlo to price the continuous downside, and a decision tree wherever a discrete event or a real option sits at the center of the thesis.
Scenario comparison across sites is a modeling exercise too. For a new facility, the question is rarely “does this location work?” but “does it work better than the alternatives?” Wastenaut’s scenario comparison workflow lets you run the same model across candidate sites, each with its own fee environment, feedstock radius, and regulatory context, side by side against the same underlying data. Pair it with Nexus facility data so every case reflects the actual market rather than an assumption.
The Part That Actually Matters: Grounded Inputs
Here is the uncomfortable truth about all three techniques: the mechanics are the easy part. A what-if table, a Monte Carlo engine, and a decision tree can all be built in a spreadsheet by an analyst in a day. What separates a scenario model that protects capital from one that produces a precise-looking illusion is entirely the quality of the inputs.
Consider what each technique demands. What-if analysis needs a credible worst case for each variable, not a round-number guess. Monte Carlo needs a distribution for each uncertain input, its center and, just as important, its realistic spread and its correlation with the others. A decision tree needs probabilities for discrete events and payoffs for each branch. Every one of those is an empirical question about a specific market, and every one is where the analysis quietly goes wrong when the input is intuition dressed up as data.
Where do grounded inputs come from? Tipping-fee ranges come from historical and current fee data for the specific region, not a national average. Feedstock and volume distributions come from generator-level supply data, contract history, and utilization rates at comparable facilities. Regulatory probabilities come from tracking the actual legislative and rulemaking pipeline in the relevant jurisdiction, the bans debated, the mandates phased in, the surcharges proposed. None of these is available from a single consultant’s point estimate or a memory of what fees “usually” do.
This is the case for building the scenario model on top of real facility-level data rather than assumption. Wastenaut maps facility-level tipping fees, capacity, and feedstock generators across the US so the ranges, distributions, and probabilities in your model reflect the market that exists. You can benchmark those inputs directly in Nexus, verify a specific load-bearing assumption against independent data before you rely on it through Validate, or scope the competitive and regulatory context of a whole market with a structured market survey. The scenario technique is only as honest as the numbers feeding it, and the numbers are where waste-specific data does the work that generic finance content cannot.
How Scenario Modeling Fits the Broader Underwriting Process
Scenario modeling does not stand alone. It sits inside a sequence that starts with a base financial model and ends with a diligence decision, and understanding where it fits keeps you from asking it to do a job it cannot.
The foundation is the base-case model itself. Before you can vary an assumption, you need a clean cash-flow model of the facility: revenue, operating cost, capital expenditure, financing, and the return metric you care about. That is the subject of building a waste facility financial model, and every scenario technique in this article is a layer applied on top of it. A scenario model built on a shaky base model inherits every one of the base model’s errors and adds a false sense of rigor.
From the base model, sensitivity analysis is often the connective step between the model and full scenario work. Sensitivity analysis systematically ranks which inputs move the outcome most, which tells you where to focus the more expensive scenario techniques. It is the bridge from “here is a return” to “here are the two or three assumptions that determine whether that return is real.”
And scenario modeling feeds diligence rather than replacing it. Every technique here produces a list of load-bearing assumptions, the facts the deal depends on, but none of them verifies those facts. That is the handoff to due diligence: the scenario model tells you which assumptions to confirm, and diligence confirms them against independent evidence. A scenario pass that surfaces a critical assumption and then does nothing to verify or structure around it is wasted effort. Used in sequence, base model to sensitivity analysis to scenario modeling to diligence, the whole process turns a single-point forecast into a defensible investment decision.
Frequently Asked Questions
What is scenario modeling in waste infrastructure investing?
Scenario modeling is the practice of representing possible futures for a waste facility investment and measuring their impact on the return, rather than relying on a single-point forecast. In practice it means taking the assumptions a deal rests on, above all the tipping fee, the feedstock or throughput volume, and the regulatory environment, and testing how the investment performs when they move. It is an umbrella that covers several techniques, principally what-if analysis, Monte Carlo simulation, and decision trees, each suited to a different kind of uncertainty.
What is the difference between what-if analysis, Monte Carlo simulation, and decision trees?
They answer different questions. What-if analysis changes one variable at a time to a specific alternative value and shows which assumptions are load-bearing; it is fast and simple but cannot capture variables moving together. Monte Carlo simulation varies all uncertain inputs simultaneously according to probability distributions and returns the full distribution of outcomes, including the probability of loss; it handles continuous, interacting risk. Decision trees model discrete events, such as a permit or a regulatory vote, and sequenced or optional choices, rolling probabilities and payoffs back to an expected value. Disciplined underwriting frequently uses all three in sequence.
How do you model regulatory risk in a waste deal?
Regulatory changes, organics bans, diversion mandates, per-ton surcharges, and extended producer responsibility laws, are usually discrete events with a probability and a timeline, which makes them well suited to decision-tree modeling and to directional what-if cases. The key is to model the specific, foreseeable changes in the relevant jurisdiction rather than a generic haircut: identify the mandate or surcharge actually in the legislative or rulemaking pipeline, estimate its probability and its effect on your tonnage or cost basis, and test whether the deal still clears its hurdle if it lands. Because regulation is often the most foreseeable of the major risks, it is also the most damaging to leave unmodeled.
Why does waste-specific scenario modeling differ from generic finance scenario modeling?
Generic scenario-modeling guidance varies abstract categories like revenue and cost, which is too coarse to be actionable for waste. Waste deals break on three specific variables, tipping-fee volatility, regulatory change, and feedstock or volume supply risk, that behave in particular ways and tend to move together in a soft market. A model that does not center these, and that does not ground their ranges in real facility-level data for the specific market, is modeling the wrong thing with false precision. The techniques are shared with general finance; the variables, the correlations, and the data sources are not.
What data do you need to build a credible waste scenario model?
Grounded inputs, not intuition. You need tipping-fee ranges from historical and current data for the specific region, feedstock and throughput distributions from generator-level supply data and comparable-facility utilization rates, and regulatory probabilities from the actual legislative and rulemaking pipeline in the relevant jurisdiction. Every scenario technique inherits the quality of these inputs directly: a distribution or probability that is really a guess produces a precise-looking guess. This is why the analysis should sit on top of verified, facility-level market data rather than national averages or a single point estimate.
See Your Deal’s Real Risk Before the Money Moves
Scenario modeling is the difference between a clean IRR that looks defensible until the first assumption misses and an honest view of what a waste deal can actually do. The techniques are learnable in an afternoon; the edge is in feeding them real market data. Wastenaut maps facility-level tipping fees, capacity, feedstock, and market structure across the US so every case in your model reflects the market that exists, not the one your pro forma assumed.
Start a free demo and put your next deal’s assumptions to the test. The first month is free, and no payment method is required to begin.