[Sep 2025]Mempool empty, Consolidate your small inputs @0.15 sat/vbyte

[Pmalek]

The sad part: almost all low-fee transactions are just data spam abusing the Bitcoin blockchain to earn money from gullible people. I'm really surprised barely anyone uses Bitcoin for normal payments nowadays.

It is sad how little normal payments activity there currently is on the Bitcoin blockchain. At the same time, if we didn't have the spam we all hate, most blocks would be mined empty and miners would earn even less than they do today, not counting the standard block subsidy. The spam at least makes the network look busier until your scratch the surface and look closer at what is really happening.

[LoyceV]

if we didn't have the spam we all hate, most blocks would be mined empty and miners would earn even less than they do today

You make it sound as if miners don't earn a lot. They earn over 50 million dollars per day now, and a large share of those Bitcoins must be sold for dollars to pay for hardware and electricity. If they earn more, all they'll do is buy more hardware to increase their share and increase the hash rate, but more money leaves the Bitcoin ecosystem.

[Pmalek]

You make it sound as if miners don't earn a lot. They earn over 50 million dollars per day now, and a large share of those Bitcoins must be sold for dollars to pay for hardware and electricity. If they earn more, all they'll do is buy more hardware to increase their share and increase the hash rate, but more money leaves the Bitcoin ecosystem.

No, no, I know they earn a lot. Especially since there are 5-10 big mining companies that mine 99% of the Bitcoin blocks. Some more, some less. Small mining operations and solo miners only hit the jackpot occasionally. I was only commenting on the fees miners earn from the transactions that are mostly spam. The profit from the fees is down and it would be even lower if it wasn't for the spam (wanted or unwanted depending on who you ask). We know that the subsidy rewards will be halved many more times in the future. But as Bitcoin's value continues to go up, those sats are also worth more in fiat terms even if there are fewer BTC units.

[stwenhao]

I'm really surprised barely anyone uses Bitcoin for normal payments nowadays.

And I'm not surprised at all. When a lot of users are pushed into centralized services, and many of them transact on second layers, then the consequences are simple: you have less and less on-chain users.

Also, when I asked Garlo Nicon to send me some coins from Revolut, then guess what: they blocked him, because of suspicious activity. So, if only transactions between exchanges are considered "safe", and if there are a lot of parties, handling coins for users, then the end result is not difficult to predict.

Currently, regulations are slowly killing Bitcoin. And many people are switching to other things, because who cares, if one BTC will be worth millions, if you won't be able to legally use it?

Also, if someone is going to trade, then it can be done, without ever touching any coins. There are CFDs, and other instruments, and quite often, you can trade, while having only fiats.

[joker_josue]

The sad part: almost all low-fee transactions are just data spam abusing the Bitcoin blockchain to earn money from gullible people. I'm really surprised barely anyone uses Bitcoin for normal payments nowadays.

Unfortunately this is happening some time ago.

People have become afraid to buy things with BTC.
They get the idea that this compromises their privacy.
But it will be very difficult to make online purchases, using BTC, without being able to share some information (especially if it is a physical good)

Where you were available to use BTC for purchases
What would give you the most confidence to use?

[LoyceV]

The profit from the fees is down and it would be even lower if it wasn't for the spam (wanted or unwanted depending on who you ask).

This made me curious about the long-term fee development: let's check the data: Bitcoin block data available in CSV format (fee_total_usd). This graph shows the full Bitcoin transaction fee history in dollar per block:

(I reduced the vertical scale to remove outliers with ridiculous fees, to make the graph readable)

As a rough estimate from this graph, I'd say transaction fees per block (measured in dollars, not Bitcoin) have been slowly increasing since the beginning because of the increase in Bitcoin value.

We know that the subsidy rewards will be halved many more times in the future.

As I read years ago: it's several halvings too early to start competing on transaction fees.

Handling this data in a spreadsheet is terrible: it's unresponsive and consumes loads of RAM. This made my computer unresponsive for a few hours! Remind me never to do this again....

[Pmalek]

This made me curious about the long-term fee development: let's check the data...

As a rough estimate from this graph, I'd say transaction fees per block (measured in dollars, not Bitcoin) have been slowly increasing since the beginning because of the increase in Bitcoin value.

That's also what I said. Block subsidy aside, miners are earning less BTC/sats than they used to. Even before 0.1 sat/vByte became the new standard, there was very little congestion on the Bitcoin blockchain despite BTC hitting new ATHs several times. On top of that, the block subsidy has been halved a couple of times already. However, those sats are worth more in USD or any other fiat terms. It would be interesting to see a comparison against gold.

Btw, does this graph go all the way back to the beginnings of Bitcoin in 2009?

[LoyceV]

That's also what I said. Block subsidy aside, miners are earning less BTC/sats than they used to. Even before 0.1 sat/vByte became the new standard, there was very little congestion on the Bitcoin blockchain despite BTC hitting new ATHs several times. On top of that, the block subsidy has been halved a couple of times already. However, those sats are worth more in USD or any other fiat terms. It would be interesting to see a comparison against gold.

I'm not making another graph against gold But you have a point, my "rough estimate" will probably change a lot if you account for inflation in this graph.

Btw, does this graph go all the way back to the beginnings of Bitcoin in 2009?

Yes: vertical are dollars, horizontal are blocks. It starts from block 0. I would have added a label, but that would have taken a few more hours....

[stompix]

They earn over 50 million dollars per day now, and a large share of those Bitcoins must be sold for dollars to pay for hardware and electricity. If they earn more, all they'll do is buy more hardware to increase their share and increase the hash rate, but more money leaves the Bitcoin ecosystem.

And more electricity is spent on mining and the more money is spent on additional hardware, the more secure the network is. Have the reward go down to 10 million and any rogue entity would be able to attack the network at a fraction of the cost now by buying gear at scrap metal prices 

As a rough estimate from this graph, I'd say transaction fees per block (measured in dollars, not Bitcoin) have been slowly increasing since the beginning because of the increase in Bitcoin value.

I just checked out of curiosity a few blocks, it seems a bit weird but reasonable once you think about:
https://mempool.space/block/0000000000000000000c17ceb6bdf762bfffe1a744b32d164c1b49b0d682a0af
half-empty blocks fees were

Total fees   ‎0.066 BTC $556.59

(yes it's price adjusted)

It is sad how little normal payments activity there currently is on the Bitcoin blockchain. At the same time, if we didn't have the spam we all hate, most blocks would be mined empty and miners would earn even less than they do today, not counting the standard block subsidy.

The fees from under 1sat/vb tx are negligible, a speck compared to the total reward, not worth it in some cases
https://asktom.cf/index.php?topic=5237323.msg65704605#msg65704605
 

[Pmalek]

I'm not making another graph against gold

I understand that.

The fees from under 1sat/vb tx are negligible, a speck compared to the total reward, not worth it in some cases

What are you comparing the below 1 sat/vByte fees with? With transactions with fees that pay 1 sat/vByte and more or with the block subsidy? If it's with the former, then I agree with you. If it's with the latter, then I don't. As we speak, Jochen Hoenicke's mempool shows 48 vMB of unconfirmed transactions. 0.8 vMB are transactions paying +1 sat/vByte. The rest, which is a huge majority, pays less than that.

Miners don't earn much from the transaction fees alone, especially if compared with the block subsidy. But if we only look at the fees, it's the sub 1 sat/vByte transactions that take up the most space in blocks and those transactions earn miners most of the money they get from blockchain fees.

[joker_josue]

They earn over 50 million dollars per day now, and a large share of those Bitcoins must be sold for dollars to pay for hardware and electricity. If they earn more, all they'll do is buy more hardware to increase their share and increase the hash rate, but more money leaves the Bitcoin ecosystem.

And more electricity is spent on mining and the more money is spent on additional hardware, the more secure the network is. Have the reward go down to 10 million and any rogue entity would be able to attack the network at a fraction of the cost now by buying gear at scrap metal prices 

Not necessarily. The network can continue exactly like this, without growing, and will remain profitable and secure.

The problem, so to speak, isn't security, but rather that large miners are businesses. Like any business, they have to seek to increase their profits year after year. The best way to do this is by acquiring more hash power.

But this isn't necessary to make the network more secure. It's just necessary to make them even more money.

[stompix]

What are you comparing the below 1 sat/vByte fees with? With transactions with fees that pay 1 sat/vByte and more or with the block subsidy? If it's with the former, then I agree with you. If it's with the latter, then I don't.

It's not me it's CK:

Now that solo ckpool has mined a block with sub 1sat/vB transactions I have data on the reward mined as the result of a low transaction fee period. Of the ~4900 transactions mined in this block, ~3300 were sub 1sat/vB transactions. It took some time to sift through the data of these transactions to determine what the mined fees were. They amounted to ~.0018 BTC more in fees, or ~$220. This is ~.06% of the current 3.125 BTC block reward.

Whole block in question fees were ~‎0.012 BTC $1,455, so ~ 13% of total fees.

Not necessarily. The network can continue exactly like this, without growing, and will remain profitable and secure.

Well, you can have a
- 2 trillion worth of coins chain protected by a gear that needs 50 million a day to be overpowered
- a 10 trillion worth of coins that is protected by gear needing 60 million a day to be overpowered
Which one feels safer?

I said it many times, you can guard an ice cream shop with a guy, you can guard a Walmart with two armed bodyguards, but I doubt guarding the White House with 3 guys is a good idea.

Anyhow, one thing I'm curious about this mountain of consolidations tx, how many of those are not actual exchnges,casinos,mixers but the same guys that flood the chain with runes?

[joker_josue]

Not necessarily. The network can continue exactly like this, without growing, and will remain profitable and secure.

Well, you can have a
- 2 trillion worth of coins chain protected by a gear that needs 50 million a day to be overpowered
- a 10 trillion worth of coins that is protected by gear needing 60 million a day to be overpowered
Which one feels safer?

I said it many times, you can guard an ice cream shop with a guy, you can guard a Walmart with two armed bodyguards, but I doubt guarding the White House with 3 guys is a good idea.

They are different things!
The network is secure due to the decentralization of the hash, not its quantity. Of course, if it's cheap to buy a lot of hash, that complicates things, but assembling hash isn't cheap.

In other words, having 5PH is as secure as having 1PH. It's not secure for someone to have +50% of the network's hashrate.

What makes it more insecure is knowing that only 3 mining pools concentrate +50% of the hash. Considering that the three largest pools gather +50% of hash, the largest pool being private and no one knows how many individual miners it has, and knowing that the second largest belongs to the largest ASIC manufacturer, I see this as more dangerous.

This idea that more hashing is needed to secure the network is an exaggeration and just a tool to feed hardware manufacturers.

[Cricktor]

Handling this data in a spreadsheet is terrible: it's unresponsive and consumes loads of RAM. This made my computer unresponsive for a few hours! Remind me never to do this again....

Have you ever tried Gnuplot? From own experience, it handles large datasets to produce a graph quite nicely. I didn't look at the original data and therefore can't say what amount of "data massage" is needed to get the correct data to plot.

[stwenhao]

I said it many times, you can guard an ice cream shop with a guy, you can guard a Walmart with two armed bodyguards, but I doubt guarding the White House with 3 guys is a good idea.

If you want to be protected by more Proof of Work, then say it explicitly inside your script:

Code:

OP_SIZE <difficulty> OP_LESSTHAN OP_VERIFY
OP_CODESEPARATOR
<pubkey> OP_CHECKSIG

Or even, you can make it signable in a very similar way, as existing P2WPKH outputs are:

Code:

OP_SWAP OP_SIZE <difficulty> OP_LESSTHAN OP_VERIFY OP_SWAP
OP_CODESEPARATOR
OP_DUP OP_HASH160 <pubkeyHash> OP_EQUALVERIFY OP_CHECKSIG

And then, it is up to you, how much Proof of Work you want to require in your transactions. Then, even if the main chain is protected by regtest difficulty, double-spending your transaction can be very difficult, and require a lot of Proof of Work.

For example: try double-spending this transaction: https://mempool.space/tx/8349df0753e80cce322322f1b76789e1d0fd6693aed2f4de4e49576423081ae7

See? It is difficult. First, because it is covered by mainchain Proof of Work, and second, because even if it wouldn't be, then you would still need a few days of grinding with GPUs, to even produce an alternative valid version of that transaction, and even move the same coins in regtest.

So, it is up to users, how much Proof of Work they will want to spend in the future, to protect their own coins.

[Pmalek]

It's not me it's CK:

Now that solo ckpool has mined a block with sub 1sat/vB transactions I have data on the reward mined as the result of a low transaction fee period. Of the ~4900 transactions mined in this block, ~3300 were sub 1sat/vB transactions. It took some time to sift through the data of these transactions to determine what the mined fees were. They amounted to ~.0018 BTC more in fees, or ~$220. This is ~.06% of the current 3.125 BTC block reward.

Whole block in question fees were ~‎0.012 BTC $1,455, so ~ 13% of total fees.

CK compared the transaction fees earned from the sub 1 sat/vByte transactions in their block with the block subsidy of 3.125 BTC. It would be better to compare transaction fees earned from sub 1 sat/vByte transactions with transactions paying +1 sat/vByte. That's a better way because it shows how much you earn from those two individual groups. Based on the numbers CK posted, their block was filled to nearly 70% with transactions paying below 1 sat/vByte. Probably that's where most of the fees came from as well.

But even if that wasn't the case. Why leave $200 worth of BTC in the air and not pocket it when we know where bitcoin is heading in the long-term?   

[joker_josue]

But even if that wasn't the case. Why leave $200 worth of BTC in the air and not pocket it when we know where bitcoin is heading in the long-term?   

For greed! This has nothing to do with network security.
This is guys who even if the sat was worth $1000, would say that the minimum value of a transaction should be 1 sat.

Hopefully I mine a block and earn $200 in my wallet. That $200 alone will cover the energy used to mine the block (maybe it's even too much).

But, if you want, I can give you my BTC address, and transfer all the fees to me, for transactions below 1sat/vb. 

[LoyceV]

Why leave $200 worth of BTC in the air and not pocket it when we know where bitcoin is heading in the long-term?

I was indeed surprised to read his explanation:

In light of how ridiculously small the extra fees mined were by accepting these transactions, and the potential for creating a significant number of new UTXOs, I am reversing my decision to mine these transactions.

So he's saying it's worth $220 to keep Bitcoin's blockchain about 1 MB smaller (until another miner includes them anyway).

But, if you want, I can give you my BTC address, and transfer all the fees to me, for transactions below 1sat/vb. 

I was hoping to get dust change sent as donation to my Electrum server, but somehow the donation address feature in Electrum doesn't work with Fulcrum server.

[stwenhao]

But, if you want, I can give you my BTC address, and transfer all the fees to me, for transactions below 1sat/vb.

How much power do you have? Enough to get 53k sats?

To solve it, you need to mine a transaction, where DER signature would have less than 53 bytes. Which in practice means grinding around 2^64 hashes, if you have nothing better, than half of the generator, as your r-value.

The current block reward is somewhere around 3.16108918 BTC per 0x170211ac target. Let's count the work per satoshi in a linear way:

Code:

0000000000000000000211ac0000000000000000000000000000000000000000 per 316108918 satoshis
0000000000000000000211ac0000000000000000000000000000000000000000 per 0x12d77076 satoshis
00000000000026fbd82365480000000000000000000000000000000000000000 per 1 satoshi
000000000000000abbac40000000000000000000000000000000000000000000 per 53000 satoshis

Which means, that it is still more profitable, to mine Bitcoin directly, than to solve the next address from my puzzle. However, coins can be added, as demonstrated here: https://mempool.space/tx/8349df0753e80cce322322f1b76789e1d0fd6693aed2f4de4e49576423081ae7 (and the fact of bumping the reward doesn't have to be public; each sponsor can fund a different miner, if needed)

And then, it is all about incentives. Users can send their coins to addresses like that, and then, miners could claim them, by grinding signatures.

it's worth $220 to keep Bitcoin's blockchain about 1 MB smaller (until another miner includes them anyway)

Of course. Signet blocks now have 1 MB limit, instead of 4 MB. Mainnet mining pools could also produce smaller blocks, if they would want. Which means, that when luke-jr talked about smaller blocks, then he could enforce it in practice, in blocks, produced by his favourite mining pools. Of course, other miners could still make bigger blocks, but still: as long as blocks are not reorged, then if one group of miners is mining less filled blocks than others, then they contibute to the smaller total blockchain size.

By the way: you got 40 sats, where you wanted, good luck sweeping them: https://mempool.space/address/bc1qzeflmklr2l4kdjkx4ak027jcg8jcs3hfne4fqy

I guess you could use the trick with negative fees, to make sweeping it profitable, send that PSBT as an invoice, then receive a signed transaction, with normal fees (paid by someone else), and then share it with miners, if you want to see it cleared. Or wait years, maybe decades, when that amount will be more significant, and when dust limits will be lifted, when miners will stop enforcing them, and Core will decide to lift yet more limits, because of that, just like they lifted OP_RETURN limit.

[Pmalek]

For greed! This has nothing to do with network security.
This is guys who even if the sat was worth $1000, would say that the minimum value of a transaction should be 1 sat.

If it's greed then they have a weird way to display it. OK, they want the minimum fee for transactions to be accepted in blocks to be 1 sat/vByte. Fine. However, others obviously don't agree and that's why they decided to accept transactions paying much less. The best they can do now is take what they can, be it an $1000 or $200 worth of BTC.