Uptime Math: How 1% Downtime Eats ROI
Every hour your miner sits offline, the network keeps producing blocks without you. That lost time does not roll over, it does not compound back, and it does not care about your electricity bill.
By the end of this article you will be able to calculate exactly how much downtime is costing you in lost revenue, identify the most common causes of unplanned stoppages, and take practical steps to close the uptime gap before it pushes your breakeven date out of reach.
Note for South Africa:
- Load shedding is the single biggest unplanned downtime risk for South African operators. At Stage 2, expect roughly 4 hours of outages per day. At Stage 6, that figure can reach 12 hours or more in some areas, according to Eskom's official load shedding stage definitions.
- Revenue is earned in USD-denominated crypto while costs are paid in ZAR. A weak rand amplifies revenue in local terms, but it also means your cost-per-kilowatt-hour feels more painful when Eskom tariffs increase.
- Prepaid and time-of-use tariffs offered by some municipalities add a variable cost layer. Use your own current tariff figure in any ROI calculation, not a number you read online.
At a glance:
- 1% annual downtime equals roughly 3.65 days of lost production across a full year.
- The financial cost of that downtime depends on your hashrate, your hardware model, and the current hashprice at the time of the outage.
- South African operators face structural downtime risk from load shedding that is not present in most other mining markets.
- Monitoring tools like Minerstat and HiveOS can track per-device uptime and alert you within minutes of a crash.
Key takeaways:
- Effective hashrate, not rated hashrate, is what actually earns revenue. Downtime is the gap between the two.
- Downtime during a low-difficulty window costs more in relative terms than the same outage during a high-difficulty period.
- A UPS, inverter, or solar-hybrid setup is the most practical uptime protection for SA operators facing load shedding.
Why Uptime Is a Hidden Profit Lever
Most operators spend time optimising overclock profiles, switching pools, or chasing marginally better electricity rates. Uptime rarely gets the same attention, yet it is the single variable with the most direct and linear impact on revenue. If your miner is offline, your revenue is exactly zero for that period. No other variable produces that outcome so cleanly.
The reason uptime gets overlooked is that small losses feel invisible. A 15-minute crash at 2am does not appear in your monthly bank statement as a line item. But those events accumulate. Over a full year, even a modest 1% average downtime rate translates to approximately 3.65 days of zero production. For a machine running near the edge of profitability, that gap can be the difference between breaking even and operating at a loss.
The Difference Between Hashrate and Effective Hashrate
Your miner has a rated hashrate printed on the spec sheet. That is the number it produces when it is running, stable, and hashing correctly. Effective hashrate is the average output your machine actually delivered across a measurement period, including all offline time, crash recoveries, thermal throttle events, and pool disconnections.
As Braiins explains when covering how mining ROI is calculated, it is effective hashrate that determines your pool earnings, not the number on the box. A 200 TH/s miner running at 98% uptime delivers an effective 196 TH/s. A machine running at 90% uptime delivers only 180 TH/s. The gap is permanent and unrecoverable.
The Uptime Math: Breaking Down the 1% Rule
One percent sounds small. In practice, 1% of a full year is 87.6 hours. That is more than three and a half days of zero production. Across a fleet of even five machines, the multiplier becomes significant very quickly.
How to Calculate Daily, Monthly, and Annual Revenue Loss from Downtime
The cleanest unit for this calculation is hashprice, which Luxor defines as the expected daily revenue per terahash per second of mining power. You can check the current figure using Luxor's hashprice explainer or track it live on Hashrate Index. Hashprice fluctuates daily with coin price and network difficulty, so always use the current or period-average figure for your calculations.
Use the step-by-step guide below to calculate your own downtime cost. Do not use figures from this article for coin price, tariff rates, or hashprice. Plug in your own current numbers at each step.
- Record total hours in the period. For one month, use 720 hours. For one year, use 8,760 hours.
- Record actual uptime hours. Pull this from your pool dashboard, monitoring software, or device logs. If you do not have logs, estimate conservatively.
- Calculate your uptime percentage. Divide actual uptime hours by total hours, then multiply by 100. Example: 700 uptime hours divided by 720 total hours equals 97.2% uptime.
- Find your effective daily revenue. Multiply your miner's rated hashrate (in TH/s) by the current hashprice (in USD per TH/s per day). This is your daily revenue at full uptime.
- Calculate lost hours and lost revenue. Subtract actual uptime hours from total hours to get downtime hours. Divide downtime hours by 24 to get downtime days. Multiply downtime days by your daily revenue figure to get your total lost revenue for the period.
- Compare against monthly operating costs. Divide your monthly electricity and overhead costs by 30 to get a daily cost figure. Add lost revenue to your daily cost and project forward to see how far your breakeven date has shifted.
This framework works for any miner model and any coin. The inputs change, the method does not.
Compounding Losses: Downtime During Peak Profitability Windows
Not all downtime hours cost the same. Bitcoin network difficulty data shows that hashprice drops when difficulty rises and climbs when difficulty falls. An outage during a post-adjustment low-difficulty window costs more in absolute terms than the same outage during a peak-difficulty period. If your miner is offline during a favourable window, the loss is disproportionate to the raw hours missed.
South African operators face a version of this problem during load shedding. If Stage 4 or higher is implemented during a period of low network difficulty, the overlap of forced downtime and high hashprice is a worst-case scenario for ROI.
Downtime Causes by Type and Severity
| Cause | Type | Recovery Time | SA-Specific Risk |
|---|---|---|---|
| Load shedding (Eskom) | Power interruption | Minutes to hours | High – structural and recurring |
| Hardware fault (PSU, hashboard) | Component failure | Hours to days | Medium – PSUs vulnerable to surge events |
| Thermal throttling | Partial performance loss | Ongoing until resolved | Medium – summer heat and poor ventilation |
| Crash loop or firmware hang | Software fault | Minutes if auto-restart is set | Low to medium |
| Network or pool disconnect | Connectivity issue | Minutes | Low to medium |
| Manual maintenance | Planned downtime | Controlled | Low – planned, not lost |
Hardware Faults, Thermal Throttling, and Crash Loops
ASIC miners run hot and they run continuously. Hashboard failures, fan degradation, and PSU stress are common causes of unplanned downtime, particularly on older hardware. Thermal throttling is a partial failure mode: the miner stays on but reduces output to protect itself from heat damage. This counts as partial downtime in any honest ROI calculation because you are earning less than rated hashrate.
Crash loops happen when a miner gets stuck in a reboot cycle, often after a firmware issue or a corrupt configuration. Without a watchdog or monitoring agent, a machine in a crash loop can sit in that state for hours before anyone notices. If your hardware is showing repeated fault patterns, it may be time to consider a professional assessment. Our professional repair and services team works with mining hardware and can help diagnose recurring faults.
Power Interruptions and the South African Load Shedding Factor
For South African operators, load shedding is not a rare event. It is a recurring, scheduled, and sometimes unscheduled structural risk. According to MyBroadband's load shedding hours tracker, cumulative outage hours in South Africa have run into the thousands annually in recent years. An unprotected mining rig takes the full impact of every outage.
The broader economic cost of load shedding across South African industries has been extensively documented, with estimates of the national productivity loss running into hundreds of millions of rand per day. For a mining operator, the per-machine cost is smaller but the proportional impact on ROI can be severe. A rig already operating near breakeven can be pushed into loss territory by load shedding alone.
Beyond direct lost production time, repeated power cycling from outages puts stress on PSUs and control boards. Secondary hardware failures driven by frequent power-on and power-off events are a real and documented risk. Factor potential repair or replacement costs into your downtime cost model, not just the lost hashing hours.
Measuring Your Real Uptime: Tools and Methods
If you are not actively measuring uptime, you are guessing. Pool dashboards give a partial picture. Per-worker hashrate graphs on platforms like Braiins Pool can reveal offline gaps, but they do not always distinguish between a crashed miner and a network drop. Dedicated monitoring software closes that gap.
- Minerstat: Tracks per-device uptime percentage, sends alerts on hashrate drops or device offline events, and supports remote restart. Works with both GPU rigs and ASICs.
- HiveOS: The watchdog feature triggers an automatic reboot if hashrate falls below a configured threshold. Downtime logs are available for review and uptime percentage can be calculated from the data. Good option for GPU rig operators.
- Pool dashboards (Braiins, Antpool, Luxor): Show per-worker hashrate over time. Gaps in the chart are visual indicators of downtime. Not as granular as dedicated software but useful for a quick check.
- Manual log sheets: A simple spreadsheet recording daily start and end times and any noted interruptions is better than nothing. It becomes the baseline for your uptime calculation.
The goal is to generate a number: uptime percentage over a defined period. Without that number, you cannot know how much revenue you are leaving on the table.
Uptime Targets: What Operators Should Actually Aim For
Commercial mining hosting providers typically specify uptime guarantees of 95% to 99% in their SLA agreements, as noted in Compass Mining's guide to mining uptime. For a self-operated machine in a home or small farm environment, 95% is a realistic floor. Below 90%, you need to treat the uptime problem as a primary business risk, not a maintenance footnote.
In South Africa, an operator without any power backup running during sustained Stage 4 load shedding may find their effective uptime already below 90% from load shedding alone, before accounting for any hardware or software faults.
Common Mistakes
- Using rated hashrate instead of effective hashrate in ROI projections, which makes performance look better than it is.
- Ignoring partial downtime events like thermal throttling because the machine "is still on".
- Not logging or tracking downtime, which makes it impossible to calculate the true cost.
- Assuming all downtime hours cost the same without checking where hashprice was at the time of the outage.
- Factoring in electricity cost but not the cost of hardware stress and secondary failures caused by load shedding power cycling.
- Setting up a monitoring tool but not configuring alerts, so the data exists but no one acts on it.
Practical Steps to Recover Lost Uptime and Protect ROI
Once you know your current uptime percentage and the revenue it is costing you, the next step is prioritisation. Not every fix costs the same, and not every improvement has the same return. Start with the highest-impact changes first.
If you are new to tracking uptime:
- Install a monitoring tool (Minerstat or HiveOS) and let it run for a full month before drawing conclusions.
- Record your first uptime percentage as a baseline. You cannot improve what you have not measured.
- Check your pool dashboard for per-worker hashrate gaps over the past 30 days to estimate your current effective uptime.
- Confirm that auto-restart or watchdog features are enabled on your monitoring software.
- Browse our mining hardware shop if ageing equipment is the root cause of frequent crashes.
If you have already tracked uptime before:
- Compare your uptime percentage against your monthly operating costs to calculate your breakeven sensitivity.
- Identify which downtime cause is the largest contributor: load shedding, hardware faults, or software crashes.
- If load shedding is the main driver, invest in a UPS, inverter, or solar-hybrid solution sized for your mining load. This is the single highest-ROI uptime improvement for most South African operators.
- If hardware faults are recurring, evaluate whether repair, refurbishment, or replacement is the better economic decision. Contact us if you need an assessment of your current hardware.
- Schedule planned maintenance during predicted low-hashprice windows to minimise the revenue cost of necessary downtime.
When Downtime Pushes Your Breakeven Date Out of Range
Your breakeven calculation assumes a certain level of consistent production. Every hour of downtime extends the timeline before your hardware costs are recovered. If your original model assumed 98% uptime and you are actually running at 88%, the gap is not trivial. Depending on your hardware cost and current hashprice, that 10-percentage-point difference can shift your breakeven date by weeks or months.
If the math shows that recurring downtime has made your current setup unviable, there are three paths: fix the uptime problem, upgrade to more reliable hardware, or exit the position. If exiting, you can sell your mining equipment to recover capital rather than continuing to operate at a loss.
Frequently asked questions
Does 1% downtime really equal 3.65 days per year?
Yes. One percent of 8,760 hours (the hours in a standard year) is 87.6 hours, which equals exactly 3.65 days. The arithmetic is straightforward. The impact on your specific operation depends on what the hashprice was during those hours.
How do I find out what my real uptime has been?
Check your pool dashboard for per-worker hashrate gaps over the past 30 days. For a more granular view, install a monitoring agent like Minerstat or HiveOS and review the uptime logs. If you have no logs at all, you will need to start tracking from today and use the first month as your baseline.
Is thermal throttling counted as downtime?
It should be counted as partial downtime in any honest ROI calculation. If your miner is producing below rated hashrate due to thermal management, the gap between rated and actual output is lost revenue. A miner that is technically "on" but running at 70% capacity because of heat is delivering 30% effective downtime in performance terms.
What is the most cost-effective uptime improvement for a South African operator?
For most South African operators, protecting against load shedding with a correctly sized UPS, inverter, or solar-hybrid setup gives the highest return on uptime investment. Hardware and software faults are secondary concerns. Solve the power continuity problem first, then address the other causes.
Should I factor in repair costs when calculating downtime losses?
Yes. If a hardware fault causes a four-hour outage and then requires a part replacement, the true cost of that downtime event includes both the lost hashing revenue and the repair cost. Power cycling from load shedding also causes PSU and component stress over time, so secondary failure costs are a real variable in the South African context.
Summary
- 1% annual downtime equals 3.65 days of zero production. It is not a rounding error.
- Use hashprice multiplied by your rated hashrate to calculate the per-hour revenue cost of any outage.
- South African operators must treat load shedding as a primary ROI risk, not a background inconvenience.
- Monitoring tools like Minerstat and HiveOS give you the data you need to know your real uptime percentage.
- If downtime has pushed your breakeven date out of reach, the decision is to fix, upgrade, or exit with capital intact.
This is educational content, not financial advice.
