Why Natural Gas Looks Cheap Until It Isn't
Natural gas prices moved from under $2 per MMBtu in early 2020 to over $40 that same year during the Texas freeze, then settled back down to single digits. That volatility isn't a one-time anomaly. It happens repeatedly and it breaks spreadsheets that were built assuming stable feedstock costs. I've seen operators lock in five-year supply contracts at $3 and still get burned by take-or-pay clauses, stranded infrastructure costs, and transport surcharges that aren't indexed the way anyone told them they would be. Volatile pricing is the first thing people forget about. Unlike coal or solar, which have more predictable cost curves, natural gas is a commodity whose spot price swings wildly with weather, geopolitical disruptions, and pipeline capacity constraints. I worked on a project in the Permian basin where we projected a 15% margin based on $2.50 gas. Three months later, the price hit $8 because of a regulatory shutdown on a critical processing plant. Our margin went negative overnight and we had to curtail production. Transportation and distribution costs compound the problem. Even when wellhead prices are low, getting gas to market requires pipelines, compression stations, and LNG terminal fees. The last mile is expensive and not everyone has access to it. Plants located far from major pipeline corridors pay significantly more just to receive fuel, and those costs don't decrease linearly with distance. Infrastructure bottlenecks can create regional price spikes that have nothing to do with supply availability.
Captive plants face a structural disadvantage during peak demand windows. When electricity prices surge during heat waves or cold snaps, natural gas-fired generation is often the marginal provider, meaning its fuel cost sets the price. But when those same events spike gas demand, the fuel cost inside the power plant also rises. You're paying more for both inputs at the same time. Margins compress or disappear entirely during the exact periods when reliability should matter most. This is the paradox of peaking plants. Stranded asset risk is another hidden cost. Natural gas infrastructure has long lifespans of 30 to 40 years, but policy shifts toward decarbonization can leave plants operating well below their designed capacity factor. I encountered a combined-cycle plant that was contracted to run at 65% utilization when it was built, but after a state mandated a renewable portfolio standard, actual utilization dropped to 28%. The debt service on the plant didn't adjust. Levelized costs per MWh doubled even though the capital was already sunk.
How Pricing Actually Works In Practice
The standard approach involves three layers: wellhead commodity cost, pipeline transportation charges, and distribution or city-gate markups. Each layer introduces friction. Wellhead prices follow regional benchmarks like Henry Hub or AECO, but the benchmark price assumes you're selling gas at a specific delivery point. Moving it anywhere else costs extra, and those transportation tariffs are often regulated in ways that don't reflect real capacity constraints. When evaluating whether to convert or continue operating a natural gas facility, the calculation isn't just fuel price versus alternative fuel price. You need to model the full cost stack including capacity charges, minimum flow requirements, interruptible service penalties, and storage injection-withdrawal costs. Most people skip the last two. I run a spreadsheet that includes hypothetical storage cycles costing $0.50 to $1.20 per MMBtu depending on location, and it changes the economics significantly for seasonal operations. The hedge strategy most companies use doesn't actually solve the problem. Financial hedging through futures contracts locks in a price, but it also caps upside and introduces basis risk. If your plant is in a region where the local index diverges from the futures curve, you're hedging the wrong price. I learned this the hard way when a basis spread between our delivery point and the trading hub widened from $0.30 to $2.10 per MMBtu during a pipeline maintenance event. Our hedge covered Henry Hub prices perfectly. It covered nothing locally.
Operational inflexibility adds another layer. Combined-cycle gas plants have minimum stable combustion levels around 40 to 50% of nameplate capacity. They can't throttle down cheaply without efficiency losses or equipment wear. When grid demand drops and the plant must run below that threshold, the heat rate degrades, meaning you burn more gas per unit of electricity produced. The effective fuel cost per MWh increases non-linearly at low loads. This matters when you're competing against renewables with near-zero marginal cost that still need dispatchable backup.
What The Numbers Miss
Externalized costs are rarely captured in a standard economic analysis. Methane leakage across the supply chain runs estimated 1.5 to 4 percent depending on the measurement methodology and region. If you value that methane at its climate impact cost, which most analyses don't, the true per-unit cost rises meaningfully. There's also the question of water usage in extraction, local air quality impacts, and the carbon intensity relative to other generation sources. These don't appear on the utility bill but they appear in regulatory fines, litigation costs, and reputational risk assessments. When I evaluate a site for gas dependency, I now track something called the delivered fully burdened cost. It includes wellhead, transport, city-gate, storage, compression, capacity reservations, and an allowance for basis risk. On a typical US industrial site, that final number can be 40 to 60 percent higher than the published benchmark price. The discrepancy is large enough to change project viability entirely. The counter-intuitive part is that cheap gas doesn't always mean cheap energy for the end user. Industrial users connected to distribution networks with peaking tolls often pay more per MMBtu than a residential customer in a competitive supply zone. The infrastructure you're connected to and the tariff structure you inherited matter more than the underlying commodity price. Switching suppliers only helps if your pipeline path and capacity reservations are favorable. They rarely are for older facilities.
If you're working with tight margins on gas-dependent operations, the practical workaround is to model multiple gas price scenarios and include basis risk as a variable, not a constant. Run a sensitivity analysis at $1.50, $3.00, $5.00, and $8.00 per MMBtu delivered. Then stress-test each scenario with a plus-or-minus $1.50 basis deviation. The projects that survive that exercise are the ones worth pursuing. Everything else is a gamble.
