Hashprice at $32.86: Which Antminers Still Pay Their Power Bill, and Which Repairs Still Make Sense (November 2026)
The week ending November 2, 2026 gave fleet operators a hard number to plan around. According to the Hashrate Index weekly roundup, USD hashprice fell 8.7% over the week to $32.86 per PH/s per day, against $35.99 a week earlier. The 30-day average is $37.49. Bitcoin traded around $76,750, down 12.2% year to date. Network hashrate (7-day average) sat near 1,062 EH/s, and the last difficulty adjustment on October 29 raised difficulty 5.5% to 148.36T. The forward curve is not signalling relief: the same source puts the average hashprice implied by futures over the next six months at about $33 per PH/s per day.
For a mining farm, that is a repair-economics question before it is a market-commentary question. At this hashprice, the decision to spend money on a dead hashboard, a failed PSU or a worn fan set depends on one thing: does the machine earn more per day than its electricity costs? This article gives you the arithmetic to answer that for your own fleet, using the published figures and a calculation you can rerun yourself whenever hashprice moves.
What the market data says this week
Hashrate Index also publishes what it calls energy hashprice: the revenue a fleet earns per megawatt-hour consumed, grouped by fleet efficiency. For the same week, the figures were:
- Fleets under 14 J/TH: about $116 per MWh
- 14 to 19 J/TH: about $84 per MWh
- 19 to 25 J/TH: about $63 per MWh
- 25 to 38 J/TH: about $43 per MWh
Read those against your power price. A fleet in the 25 to 38 J/TH band earning $43 per MWh is earning 4.3 cents per kWh of revenue. Any site paying more than that for electricity is losing money on every hour that fleet runs, before cooling, hosting margin, pool fees and labour. A fleet under 14 J/TH earns 11.6 cents per kWh of revenue and has far more room.
Break-even electricity price by model
The calculation is simple. Daily revenue is hashrate (in PH/s) multiplied by hashprice. Daily energy use is power draw in kW multiplied by 24. Break-even electricity price is revenue divided by energy. The table below applies the November 2 hashprice of $32.86 to four common machines.
| Model | Hashrate / power | Efficiency | Revenue per day | Energy per day | Break-even power price |
|---|---|---|---|---|---|
| Antminer S19j Pro | 92 TH/s / 2,714 W | 29.5 J/TH | $3.02 | 65.1 kWh | $0.046/kWh |
| Antminer S21 | 200 TH/s / 3,500 W | 17.5 J/TH | $6.57 | 84.0 kWh | $0.078/kWh |
| Antminer S21 XP | 270 TH/s / 3,645 W | 13.5 J/TH | $8.87 | 87.5 kWh | $0.101/kWh |
| Antminer S23 | 318 TH/s / 3,498 W | 11.0 J/TH | $10.45 | 84.0 kWh | $0.125/kWh |
Method note: figures are our own calculation from the hashprice above and the listed specifications. They are gross of pool fees, hosting margin, cooling overhead and downtime, and they use one week's hashprice, which moves daily. Treat them as a screening tool, not a forecast.
The pattern is blunt. At roughly 7 cents per kWh, an S19j Pro loses about $1.54 per day on energy alone, an S21 earns roughly $0.69, an S21 XP about $2.75 and an S23 about $4.57. At 5 cents the S21 earns $2.37 per day; at 9 cents it loses $0.99. The S21 sits right on the line for a lot of sites, which is why its repair decisions are the most sensitive to a hashprice move of a few dollars.
The repair decision in one formula
Once you have the daily margin for a machine at your own power price, the repair question reduces to a payback calculation:
- Daily margin = revenue per day minus (kWh per day x your power price)
- Repair payback in days = total repair cost (parts + labour + downtime already incurred) divided by daily margin
- If daily margin is zero or negative, the repair does not pay back at this hashprice. The machine is worth repairing only if you expect hashprice or your power price to change.
Take an S21 XP at 7 cents. Its margin is about $2.75 per day. A hypothetical repair that costs $60 in parts pays back in roughly 22 days of operation. The same $60 on an S19j Pro at 7 cents never pays back, because the machine does not cover its power. This is why "can we fix it?" is the wrong question for a fleet in this market. The right question is "does it earn enough afterwards to justify fixing it?"
Where repairs still make sense
Three situations keep repair spending rational even at $32.86:
- Efficient machines on cheap power. An S21 XP or S23 on a low-cost site has a wide margin, and every day it sits on the bench is revenue that does not come back. For these units, the cost of waiting for a part exceeds the cost of the part.
- Cheap, high-frequency failures. Fans, connectors and cables cost little, yet when one fails it can take a hashboard, or the whole machine, offline. Replacing them is almost always the best return per dollar in the fleet.
- Machines you expect to bring back into the profitable zone. If your power contract steps down or your hosting terms change, a unit that is marginal today may be comfortably positive in six months. Repairing it now only makes sense if the repair cost is small relative to that expected margin.
Where it does not
Older hardware in the 25 to 38 J/TH band, at power prices above about 5 cents, is the clearest case. If a unit already loses money on energy, a hashboard replacement just lets it lose money for longer. These machines are better candidates for curtailment, resale for parts, or a planned retirement than for new component spend. Be honest with yourself about this: the repair budget has a better use on units that earn.
How to use this in your own fleet
- Pull the current hashprice from a source you trust and note the date.
- For each model in your fleet, compute daily revenue and daily kWh from rated hashrate and power. Use your measured power if you tune your machines.
- Apply your all-in power price and rank models by daily margin.
- Set a spending ceiling per model: the repair cost that pays back inside the window you are willing to wait.
- Rerun the sheet whenever hashprice moves by more than a few dollars, and after every difficulty adjustment. The next one is estimated around November 13, though that estimate is uncertain.
What this means for parts stock
A thin-margin market changes what you keep on the shelf. The units that matter most are the ones that earn, so the parts to hold are the ones that keep efficient machines running: cooling fans, power supplies, connectors and headers and the chips and repair components for your highest-margin models. For model-specific parts, start from the Antminer repair parts or Whatsminer repair parts collections.
If you are planning a full-year parts budget on these assumptions, our 12-month standing stock budget template shows how to size it by fleet. For the broader case on why parts cost belongs in your ROI model, see The Hidden Variable: Why Spare Parts and Repair Cost Decide Real Mining ROI in 2026. And for the question of which generation is worth repairing at all, see Antminer S21 vs S19 XP: Which One Is Worth Repairing in 2026?
FAQ
Does a low hashprice mean I should stop buying spare parts?
No, but it means you should buy selectively. Parts for machines that clear their power cost keep that margin flowing. Parts for machines that do not cover energy are rarely a good use of capital at this hashprice.
Are these break-even prices exact?
No. They are screening numbers built from rated specifications and a single week's hashprice. Your measured power draw, pool fees, hosting terms and uptime will all move the real figure. Rerun the calculation with your own inputs.
Will hashprice recover?
We do not forecast it. The futures-implied average for the next six months is close to today's level, according to the Hashrate Index data, but that is a market price, not a promise. Plan for the hashprice you have and keep your repair ceilings flexible.
Source your parts package
Tell us your fleet models, your quantity and your power price range, and we will help you decide which parts to hold and which to skip. Browse the full range in our Antminer repair parts collection, or contact us directly.
Email: contact@lys-sz.com
Direct from our warehouses in Shenzhen. Worldwide shipping. DDP available for buyers in the US and the EU. Latam, SE Asia, Middle East, CIS, Africa quoted case by case.
Market data in this article: Hashrate Index weekly roundup for the week ending November 2, 2026. Machine specifications: manufacturer-published rated figures and the S19j Pro figure cited in Luxor/CryptoSlate reporting. Calculations are ours.
