Bitcoin transactions break at scale, and miners’ incomes could collapse.
At peak capacity, the Bitcoin block size limit arbitrarily limits Bitcoin to seven transactions per second (TPS), with the common limit being closer to four TPS in practice. This bottleneck creates potentially fatal problems for Bitcoin as the user experience deteriorates when transactions are over seven TPS. Bitcoin evicts transactions from the processing queue or keeps them in a queue for all the lowest fee transactions for up to fourteen days. This has caused frustration among Bitcoin users, and reduced consumer trust. The transaction limit also prevents miners from transitioning their income from newly issued coins to transaction fees as the amount of newly issued coins is halved about every four years until all 21 million Bitcoins are issued.
Due to the seven TPS capacity, there are not enough funds to sustain the security budget of miners. In January and February of 2025, Bitcoin fees as a percentage of the block reward have ranged between 1–3%. The remaining Bitcoin miners rely on as their operating income is 97–99% newly minted Bitcoins. To put it bluntly, if this trend continues to the next halving in 2028, when the amount of Bitcoins distributed to miners is cut in half, the miners’ income will also be decreased by almost 50%, possibly leading to bankruptcies among miners.
Even with maximum adoption, a dilemma presents itself: users cannot complete their transactions without paying increasing fees and are likely to abandon Bitcoin for a more user-friendly blockchain alternative. This article further explores how at seven TPS, the fee rewards for miners are unsustainable for the budgets of the massive mining operations currently sustained mainly through block rewards. When the block rewards reach zero around 2140, will the fee market at seven TPS be enough to sustain the massive mining operation, currently serving as the backend, and the security of the Bitcoin network? Further, this article discusses how Bitcoin has faced this dilemma multiple times in recent years, and what it means for its future prospects.
Every 3–4 years the Bitcoin mining reward is cut in half. When all 21 million Bitcoins are mined, miners’ budgets will completely rely on transaction fees. Having a fixed capacity of seven TPS means that the transaction fees are unlikely to be enough to replace miners’ income from newly distributed coins.
The massive Bitcoin mining industry currently employs thousands of people and consumes around 150–170 TWh of electricity, more than the electricity consumption of Sweden, a country with over ten million inhabitants.
Minor transaction spikes break the Bitcoin user experience.
About one year before the Bitcoin block size limit was added Satoshi proclaimed that Bitcoin would scale beyond VISA. When the supposedly temporary block size limit was added limiting Bitcoin to a maximum of seven TPS capacity. Satoshi Nakamoto, Bitcoin’s founder(s) was already convinced that Bitcoin could handle a larger transaction volume than VISA. Estimating VISA to process around 15 million transactions per day, which roughly equals 174 TPS:
The existing Visa credit card network processes about 15 million Internet purchases per day worldwide. Bitcoin can already scale much larger than that with existing hardware for a fraction of the cost. It never really hits a scale ceiling. If you’re interested, I can go over the ways it would cope with extreme size. — Satoshi Nakamoto, Sun, 03 May 2009 23:32:26
Satoshi Nakamoto predicted higher capacity for Bitcoin than Visa's 174 TPS, but the block size limits Bitcoin to 7 TPS. When transaction volume exceeds the Bitcoin network capacity of seven TPS, the Bitcoin user experience breaks down, frustrating end-users and damaging consumer trust due to the extensive delays that can eventually end up with the transaction being cancelled.. Historically, adoption has soared from time to time, and user adoption leading to increased transaction volume has pushed Bitcoin to handle more than the maximum capacity of seven TPS. However, as Bitcoin is unable to handle more than seven TPS, using Bitcoin becomes an increasingly negative user experience for end-users because:
- The transaction backlog overflows. All transactions, except for the highest fee transactions, get stuck in the transaction backlog queue with an unknown waiting time, as higher-fee transactions can always be added to skip the wait. If a transaction is stuck in the backlog for over 14 days, it will be evicted from the transaction backlog, and effectively cancelled. Or if the transaction backlog is full at around 700,000 transactions in the backlog, the lowest fee transactions will be evicted from the backlog, and again cancelled.
- Transaction fees soar. When there is an ongoing pressure to transact more than Bitcoin’s capacity of seven TPS. There is competition among the highest fee transactions to be processed by the Bitcoin miners. The transaction fees increase drastically because the seven TPS picked from the growing backlog of transactions each second are the ones with the highest fees. End-users pay an ever increasing transaction fee that has been shown to surge to over $50 worth of Bitcoin per transaction.
Comparatively, VISA manages an average of 1700 TPS, with additional capacity available for transaction spikes well over 50,000 TPS. Let’s say that the incoming transaction volume of Bitcoin suddenly increases to 1000 TPS. Using that to make a rough estimate, the transaction backlog will be filled in around 12 minutes, and the lowest fee transactions will start being evicted from the queue, effectively cancelling them. The extremely high fees that come with it negatively affect the end-user experience and hurt consumer trust in Bitcoin. This has happened several times before, especially in 2017, early 2021, and early 2024 (special case with ordinals claiming more transaction space without paying extra fees).
Transaction overflow: Limited transaction queue backlog.
As mentioned above, when the backlog for unprocessed transactions fills up, transactions with the lowest fees are evicted from the backlog, which is a disastrous prospect from an end-user perspective. In some cases, the user has to wait a long time for the transaction to be processed, before the network rejects the transaction, which efficiently cancels the transaction. Having a full transaction backlog also forces Bitcoin nodes to enforce a minimum fee threshold, where transactions with fees below a minimum value are rejected outright. This ensures that when the transaction backlog is full, only transactions with a specific fee limit can enter the transaction backlog.
Let’s have a look at the historical transaction backlog technically known as the “Mempool”:
If a transaction is stuck in the backlog for more than fourteen days, it is automatically removed from the backlog to free up space. Imagine the user experience of broadcasting a transaction, then having it cancelled after one and a half weeks because it was stuck in the backlog. A terrible user experience that inevitably leads the consumer to the conclusion that Bitcoin is unreliable. As you can see in the unconfirmed transaction count graph in the image above, illustrating the amount of unprocessed Bitcoin transactions in the transaction backlog queue since 2017.
Although highly impractical, there is a workaround for this that allows you to increase the transaction fee of your pending transaction, using tools named Replace-by-fee (RBF), that lets you replace an unconfirmed Bitcoin transaction with a new one that pays a higher fee. Alternatively, using Child Pays For Parent (CPFP) you can broadcast a new transaction with a very high fee as a child transaction of the unconfirmed transaction. This will incentivise miners to choose both the low fee, and the subsequent very high fee transaction to collect its fees. There are usually no built-in ways to increase the fee of your transactions in a wallet. The end user usually has to know programming, and write a script to achieve this, which is utterly non-user friendly.
So far, the Bitcoin transaction backlog has never been full for a significant time. These are key issues and vulnerabilities to consider if the goal is to avoid a key bottleneck or if we want Bitcoin to take over for the traditional finance system such as VISA.
The first notable transaction backlog boom was during the 2017–2018 bull run. Bitcoin transaction volumes increased massively, the price of Bitcoin surged, and adoption grew. The transaction backlog frequently exceeded ⅓ of its capacity and there are reports of it spiking to full capacity, triggering low-fee based eviction of transactions from the backlog. Transaction fees spiked to $20-$50 worth of Bitcoin per transaction, and confirmation time stretched to hours and days for low-fee transactions. This risks transactions being marked as outdated and evicted from the transaction backlog.
In the 2020–2021 Bitcoin bull run, the transaction backlog again became heavily congested, as the price of Bitcoin soared, and transaction demand again increased drastically. Transaction fees rose to $10-$30, and some transactions with low fees remained unconfirmed for days, and were likely evicted from the transaction backlog.
The latest spike in unconfirmed transactions in the transaction queues occurred in early 2024, as shown in the chart. A significant contributor to this congestion is the rise of Bitcoin Ordinals, introduced in 2023, which uses a hack to enable NFTs and meme art on Bitcoin. This has generated another layer of on-chain activity. Coupled with high demand for Bitcoin transactions during a potential bull market, the transaction backlog has been pushed closer to full capacity.
Mining budget.
This screenshot is from a video tour of a new giant mining data centre in the United States. The video description says over 150 full-time workers and 200 contractors are on site. The video is from Bitcoin mining giant Riot Platform’s new Corsicana mining operation, which already has over 45,000 Bitcoin miners running with huge personnel and electricity costs. The operation needs a high income to sustain its budget. After all Bitcoins have been distributed, the block reward becomes zero in about a year (2140). The dollar value of the mining fees somehow has to cover the costs for all of this, and Bitcoin worth $2,000 — $4,000 for every block is not enough, miners are risking bankruptcy and being forced to pivot to mining a more profitable coin than Bitcoin.
When all 21 million Bitcoins are issued, how can massive mining operations sustain budgets using only transaction fees? At the time of writing, the transaction backlog in Bitcoin has cleared out, which presents another problem for miners. The fees per transaction with low Bitcoin transaction volume are extremely low. Fees per Bitcoin block start to range from $2k-$4k, which is orders of magnitude less than the income from block rewards that is currently at ~$300 000 worth of Bitcoin. This means that the majority of Bitcoin miners’ profits are based on newly distributed coins. After all 21 million Bitcoins are mined, mining revenue will no longer be based on newly minted Bitcoins based on block rewards, and thus rely solely on fees, which is unsustainable due to the 7 TPS block size limit.
There is a prevalent issue with Bitcoin blocks being full, and transaction fees being high, when the network is congested due to the block size limit. This causes the Bitcoin to process low-fee transactions. When the block rewards halve and eventually disappear, will fees from seven TPS be enough to sustain Bitcoin miners and their operations costs? Likely not, for Bitcoin to be sustainable after all 21 million coins are issued, fees must be high enough to maintain the cost of operations for miners. Otherwise, miners will lose their primary source of income.
If 5–10% of Bitcoin’s total supply were moved in a single day, assuming typical transaction sizes and distribution, a massive queue and network halt would likely occur (except for the top 7 high-fee transactions). This would generate transaction volumes far exceeding Bitcoin’s 7 TPS capacity, filling the transaction backlog, causing delays, fee spikes, and evictions of low-fee transactions.
Historically, Bitcoin fees have markedly fluctuated, as end-users start outbidding each other to process their transactions. This historically occurs when market conditions first cause the Bitcoin transaction backlog to congest, making Bitcoin’s end-user experience frustrating and unreliable.
Bitcoin miners will have to hope for significantly increased fees to make their operations profitable. However, with high fees stemming from high congestion, the user experience will be reduced, and only users who can afford high fees will be able to send Bitcoin.
Problems with soaring fees and transaction queues in the transaction backlog, with the risk of eviction likely incentivise users to move to other more user-friendly chains, with higher transaction capacity, such as Solana, Bitcoin Cash, and Nexa (This is not financial advice…). Most crypto adoption, excitement, and crypto businesses are in such alternative coins, not Bitcoin. The tools and capabilities for building products and services on top of Bitcoin are also limited. For example, it is vastly more practical for a computer science or finance student to build and use these alternative coins than Bitcoin.
