In mining, Energy Arbitrage means using electricity-price differences to decide when to mine and at what power level. The electricity is consumed in computation, not stored for later resale. This can improve operating economics, but it does not guarantee a risk-free profit.
With a battery, imagine charging during a cheap hour and supplying electricity later when it is worth more. EIA, the US Energy Information Administration, describes storage price arbitrage this way while also noting storage losses. A miner does something different: it uses the energy for computation immediately. It may earn Bitcoin revenue, but the coins are not a stock of electricity that can later be fed back into the grid. The battery analogy stops being useful when it confuses financial value with physical energy. [1]
A low exchange price may not be the price a particular miner pays. The contract may fix the price or include additional components. U.S. DOE, the US Department of Energy, distinguishes energy billed in kWh, charges for peak demand in kW, and fixed fees. A kilowatt (kW) is a thousand watts of power. Drawing one kilowatt continuously for an hour uses one kilowatt-hour (kWh) of energy. Larger sites use megawatts (MW), each equal to a thousand kilowatts, and megawatt-hours (MWh), each equal to a thousand kilowatt-hours. A short pause can reduce energy use without changing an already recorded monthly peak or a fixed fee. The calculation therefore needs the actual tariff, including relevant supply and delivery charges, rather than an attractive point on a chart. [2] [5]
Hashprice estimates revenue for a given amount of computing power over a given period; efficiency tells us how much energy that work requires. Price Adapt, a feature in the Braiins Manager mining management tool, combines these inputs. In the calculation below, USD stands for US dollars. Take our own simplified example: hashprice of 0.048 USD/TH/s/day and efficiency of 20 J/TH (joules of energy per trillion attempts). At 1 TH/s, a trillion attempts per second, over 24 hours the machine uses 0.48 kWh. Expressed per unit of energy, gross revenue is 100 USD/MWh, or dollars per megawatt-hour. We can now compare it with an electricity price in the same unit. This is not net profit: the example excludes auxiliary consumption, fees and other costs. [3]
Imagine two operating settings. The first leaves 30 USD/MWh after the electricity cost being compared, and draws 1 MW: its hourly contribution is 30 USD/h. The second has a better margin of 40 USD/MWh, but draws only 0.5 MW, giving 20 USD/h. The second makes better use of each unit of energy, yet the first contributes more per hour. Braiins’ model likewise compares hourly results. These are our illustrative numbers, not specifications for a particular machine; each setting’s revenue and power draw must be consistent with each other. [3]
According to its documentation, Price Adapt uses prices from the day-ahead electricity market (DAM) and a seven-day average hashprice. It prepares a plan that can pause and restart mining. Supported models running Braiins OS, the software controlling the miner, can also have their power target adjusted; other supported configurations may offer only pause and resume. These limits matter: the same price signal does not give every machine the same options. Nor is an average of past revenue a promise about tomorrow. Compare the planned outcome with actual operation. [3]
“Should we mine for another hour?” differs from “Will the investment pay back?” The first compares revenue with costs that change when operation continues. The second also includes machines, connection, infrastructure, servicing and expenses that continue during a pause. U.S. DOE recommends evaluating tariffs with life-cycle costs in mind. A positive result for one chosen hour cannot establish the payback of an entire farm. It is one part of the calculation, not an answer to every business question. [2]
Avoiding expensive electricity can save money compared with continuous operation. It does not automatically generate revenue from selling that electricity. U.S. DOE distinguishes variable pricing from Demand Response, contractual programs adapting demand to grid needs. Payment for reducing demand requires the specific conditions of such a program. Our comparison must be consistent: the same saving cannot be booked both as a lower cost and as new income. Deliberately limiting mining—Curtailment—describes an operating action; by itself it does not say who will pay for it. [4]
U.S. DOE recommends interval measurements to show not only how much electricity was used but when. Evaluate actual revenue and the relevant costs over the same period. Include cooling, residual consumption during pauses and the machines’ return to operation. Only then can we see the difference between the plan and the outcome. A cheap exchange hour or an attractive efficiency specification is not the result of the entire farm. Energy Arbitrage becomes useful when its economic reasoning is checked against actual operations. [2]
For the clearest picture, read this entry together with Curtailment, Hashprice, Demand Response Mining, Flare Gas Mining. The reverse links also lead from Curtailment, Flare Gas Mining, Landfill Gas Mining, Methane Mitigation.