UTXO consolidation

Fees are charged by transaction size, and size is driven by how many coins you spend. Each extra native SegWit input adds about 68 virtual bytes, which costs 136 satoshis at 2 sat/vB and 6,800 satoshis at 100. Consolidating during a quiet mempool converts an unknown future fee into a small known one today.

How it works

A consolidation is an ordinary transaction that you send to yourself, with many inputs and one output.

The arithmetic is the whole argument. Twenty native SegWit coins swept into a single output make a transaction of roughly 1,400 virtual bytes. Broadcast it on a quiet Sunday at 2 sat/vB and it costs about 2,800 satoshis. Wait until you actually need to spend and hit a busy market at 100 sat/vB, and the same twenty inputs cost about 140,000 satoshis. Fee markets do reach those levels: the halving block on April 20, 2024 collected 37.6 BTC in fees, worth more than 2.4 million dollars, as the Runes launch flooded the mempool. You cannot predict when the next crunch arrives, which is exactly why the cheap window is the one to use.

There is a ceiling. Standard relay policy caps a transaction at 400,000 weight units, or 100,000 virtual bytes, which works out to roughly 1,470 native SegWit inputs in one go. Wallets and hardware signers usually give up long before that, since signing several hundred inputs on a small device can take minutes.

The price you pay is not measured in satoshis. Spending twenty coins together publishes the claim that all twenty belong to one person, permanently. If those coins came from different places, an exchange withdrawal, a peer-to-peer purchase, a payment from a client, the consolidation ties those histories together in a single stroke. The mitigation is to consolidate in separate batches that follow the boundaries you care about, rather than sweeping everything into one output because it is tidier.

Where you see it

Small-coin accumulation is a side effect of good habits, which is why consolidation keeps coming up.

Dollar-cost averaging is the main source. A weekly recurring buy that lands in your own wallet leaves you with 52 separate coins at the end of a year, none of them large. Mining payouts, Lightning channel closures, faucet-sized gifts, and the change from every payment you make all add to the pile.

Two situations force the issue. Opening a Lightning channel needs one input large enough to fund it, so a wallet full of crumbs cannot open a channel of any size without consolidating first. Moving to a new wallet or a new hardware device means signing every coin you hold, and that migration is far less painful with ten inputs than with three hundred.

The practical routine is simple: watch the mempool, pick a moment when the next-block rate is in the low single digits of sat/vB, and sweep in batches that respect the separation you want to keep. Do not include coins you cannot explain, and do not rush a consolidation during a fee spike, because the cost is the entire reason to do it at all.

UTXO consolidation vs merge avoidance

Consolidation and merge avoidance are direct opposites and both are correct advice, for different stacks. Consolidation buys lower future fees by publishing the fact that your coins share an owner. Merge avoidance keeps that fact private and pays for it with permanently higher fees, plus the risk that small coins become uneconomic to move at all. Savings you intend to hold for years lean toward consolidation, because the fee saving compounds and the linkage is between coins you already treat as one stash. Coins you keep deliberately separate should never be swept together at any fee rate.