Abnormally high Ethereum gas fees are fees far above what a transaction’s size, the network’s conditions and the sender’s usual behaviour would explain. They are a reason to look closer, not evidence of money laundering. Because large tips are normal in some activity, teams judge a fee against the right comparison first, then look at who paid, who received it and where the funds went next.
In June 2020, someone sent 0.55 ETH and paid a fee of roughly 10,668 ETH to do it, about $2.6 million at the time. A second transfer paid the same fee the next day. The case is still the clearest example of a problem compliance teams face on Ethereum, which is what to do when a fee makes no economic sense.
The problem is harder today because the fee system has changed. Since 2021 fees have a burned base fee and an optional tip. Since the 2022 Merge, validators produce blocks, and block building is now often handled by specialised builders. Large tips are routine in some activity and a sign of error or compromise in other cases. This article explains how fees work now and what an abnormal fee looks like today. It then covers what the 2020 case still teaches and sets out a seven-step workflow for reviewing a fee without treating it as proof of crime.
Gas measures the computational work a transaction asks the network to do. The sender pays for it in ETH, and the fee is the amount of gas used multiplied by the cost per unit of gas, as ethereum.org explains. Prices are quoted in gwei, and each gwei is one-billionth of an ETH. A standard ETH transfer uses 21,000 gas units.
The value transferred does not set the fee. A 0.55 ETH transfer and a 350 ETH transfer of the same kind use the same gas, which is why a fee can exceed the amount sent. “Abnormal” therefore has to be judged against three things, namely the gas the transaction used, the prices other users were paying at the time and the sender’s own history. A large absolute fee is not abnormal if a complex contract call used a lot of gas at a normal price.
Every block carries a protocol-set base fee, and the base fee is burned. The sender can add a priority fee (a tip) that goes to the block’s fee recipient, and sets a maximum fee. The sender is refunded the difference between the maximum fee and the sum of the base fee and the tip. This came with EIP-1559 at the London upgrade, which took place on 5 August 2021 at block 12,965,000. Since the Merge on 15 September 2022, validators rather than miners propose blocks.
Two later changes matter for monitoring. Block building has been split from block proposing, which changes who actually receives tips (covered below). And since March 2024, blob transactions have a separate fee market for data posted by layer-2 networks.
In 2020 the first question was which miner received the money. Today the answer depends on how the block was built, how the tip was paid and which fee market the transaction used.
After the Merge, validators must propose valid blocks, while block builders construct profitable blocks on their behalf. Many builders and validators work through the MEV-Boost software. The block’s fee recipient field decides who receives the priority fees. Etherscan’s documentation explains that a builder can agree with a validator to be set as the block’s fee recipient. The builder then includes a transaction at the end of the block that pays ETH to the proposer’s fee recipient address.
For an analyst this means the party that received an abnormal tip may be a builder, not the staking entity behind the block. The recipient has to be identified from the block, not assumed.
Users can also pay the block creator through direct transfers to the fee recipient address made inside a transaction. A review that only reads the gas price field can miss value that reached the recipient by a direct transfer.
High payments to builders are a routine part of some activity. In competitive arbitrage, searchers often pass a large share of the value they capture on to builders and validators, through tips and direct payments. A very high priority fee or direct payment next to a trading bundle is therefore not unusual by itself. The right baseline for that transaction is the value it captured and the behaviour of similar bundles, not the typical gas price of an ordinary transfer. This is a common source of false positives for any fee-based monitoring today.
Some transactions never appear in the public mempool. One research study labels each transaction as public, private or exclusive. Exclusive transactions show up in the blocks of just one builder, while private ones reach several builders without passing through the public mempool. An analyst therefore cannot always compare a fee with what was visible in the public mempool at the time. Comparing with the fees of other included transactions in the same and nearby blocks is more reliable.
Blob transactions pay in two places, the normal gas market and the separate blob gas market. Because blobs have their own fee market independent of regular gas, an abnormal blob fee has to be compared with other blob fees, not with ordinary gas prices. Blob fees are burned like the base fee. Layer-2 networks post their data to Ethereum in blobs, but fees users pay on a layer-2 network are set by that network. Confirm which network a transaction is on before judging its fee.
The fee system has changed but expensive errors have not stopped.
In this case an unusual fee ended as a mistake, and the recipient’s return was itself evidence. Whether a recipient returns funds, and to whom, belongs in the investigation file.
The 2020 case was first reported as a 10,666 ETH fee. The more precise figure on the blockchain record is 10,668.73185 ETH. Compliance teams still study it because it shows a fee anomaly from first sighting to unresolved cause. The sections below separate what is documented from what was only reported.
On 10 June 2020, a transfer of 0.55 ETH, worth about $130 at the time, paid a fee of 10,668.73185 ETH, about $2.6 million. SparkPool processed it, and its fee was frozen promptly.
The second transfer sent 350 ETH on 11 June 2020 at 03:30 UTC in block 10241999, and Ethermine mined it. It paid 10,668.73185 ETH, the same fee as the first transfer, and was sent by the same wallet address. The transfer used the standard 21,000 gas at a price of 508,034,850 gwei (about 0.508 ETH per unit of gas). Its hash is 0xc215b9356db58ce05412439f49a842f8a3abe6c1792ff8f2c3ee425c3501023c. The screenshots show it going from 0xcdd6…182d to 0xe87f…8c30, while the first transfer went to 0x12d8…4e84.
At an illustrative 60 gwei, the same transfer would have cost about 0.00126 ETH, so the price paid was more than eight million times that level.

Ethermine’s operator, Bitfly, said on 15 June that it would share its fee among miners. It said it had given the sender a sufficient amount of time to get in contact. SparkPool, which had frozen its fee, said on 16 June that it would distribute the fee on 17 June because the sender had not contacted it. Neither announcement offered a return to the sender. The pools also received fake claims from people pretending to be the sender, and none produced a valid signature from the sending account.

Several explanations circulated. They included an honest mistake such as swapped value and fee fields, a software bug, blackmail and, more speculatively, laundering by a miner that mines its own high-fee transaction. Reports at the time did not establish the cause, the sender’s intent or who controlled the sending address.
SparkPool had faced a similar case in 2019, when a sender address paid abnormally high fees on 19 February 2019. On 20 February 2019, Scorechain advised SparkPool on social media not to return the fee, because its analysis found mixing patterns leading to the sending address. That analysis attributed about 45% of the origin of funds to ETH mixers. In March 2019, SparkPool located and verified the accidental sender of the fee and agreed to split the amount, and returned 1,050 ETH to the sender.

The outcome shows why mixer exposure raises a question for review without settling it.
It shows that the protocol lets a sender pay any amount they can afford to the party that includes their transaction, and that the payment looks like an ordinary fee. It does not show that anyone laundered money, and the sender should not be treated as a wrongdoer. It also shows what a reviewer lacks at first sight, namely the cause, the intent and the recipient’s response. Those are what the investigation workflow below is built to establish.
A fee anomaly is a prompt for review. It becomes useful to a compliance team for five reasons.
A monitoring trigger tells an analyst where to look. What the analyst finds decides whether it matters.
Most fee anomalies have ordinary explanations. The table pairs each with the circumstances that would justify looking further.
Most conventional red flags are about who is involved. They look for sanctioned or high-risk counterparties, mixing services, high-risk jurisdictions and structuring. FATF’s September 2020 report draws on more than 100 case studies and groups its indicators into six categories. They cover technological features that increase anonymity, geographical risks, transaction patterns, transaction size, sender or recipient profiles, and source of funds or wealth. FATF describes the indicators as red flags that “could suggest criminal behaviour”, not as proof. Our own overview of crypto AML covers the conventional list.
An abnormal fee is a different kind of signal. It describes economic behaviour, not identity.
The full context decides. Under a risk-based approach, a firm calibrates its response to its own risks and customers. A fee anomaly on its own is a prompt for assessment. Escalation depends on what the rest of the evidence shows.
This seven-step workflow is written for a current transaction or alert, and applies equally to a historical case.
A fee anomaly is only the starting signal. Most of the work is gathering context quickly and recording how a decision was reached. The 2020 case is used here only to illustrate the questions an analyst would ask. The Scorechain capabilities below, built on blockchain analytics, support that work.
Wallet screening. Wallet screening gives each address or entity separate risk scores for incoming and outgoing funds, on a scale of 1 to 100 where a lower score means higher risk. Scores come from entity attribution and a risk matrix based on AML and CFT standards. In the workflow above it supports steps 4 and 6.

Transaction monitoring. Transaction monitoring runs scenarios that combine transaction direction, amount and triggered risk indicators, raises alerts in real time and keeps an audit trail. For fee anomalies it helps with step 3, spotting a break from a wallet’s usual behaviour.
Risk indicators and customisable alerts. Risk indicators can be set by entity type, entity country and transaction behaviour, with limits for each indicator. Customisable alerts let teams tailor monitoring scenarios and alert rules to their business goals, operating jurisdictions and risk tolerance. In the platform view shown earlier, no risk indicator was defined for the 350 ETH transaction, and the platform offered a link to configure one. Because large tips are normal in some activity, a fee anomaly is best treated as one input among several, and teams decide which combinations of conditions deserve an alert.
Investigation and tracing. The Exploration Tool shows how funds travel between wallets, entities and transactions and presents those flows visually. Flux analysis traces assets across several wallets and paths, surfaces indirect exposure several hops away and lets users adjust how deep the analysis goes. The Case Manager supports collaborative investigations with an assigned owner, comments, files and an audit trail, and case reports can be exported. A KYT report brings a transaction’s financial data, risk scores, risk indicators and entity information into one document. These support steps 5 to 7.
It is a fee well above what the gas used, the prices other transactions were paying at the time and the sender’s own history would explain. The value transferred does not set the fee, so a small transfer can carry a very large fee. Fees are gas used multiplied by the price per unit of gas.
Most often by mistake, through a wallet setting or a software bug, or because a complex contract call used a lot of gas. In some trading activity large payments to builders are routine. Less often it follows a security compromise or a deliberate act. In the 2020 case the reason was never established.
Not by itself. A fee anomaly is a reason to review a transaction, and most have ordinary explanations. FATF describes its red-flag indicators as things that could suggest criminal behaviour, which a firm assesses under a risk-based approach using the full context.
A high fee does not by itself exempt a transaction from AML checks. A fee appears on-chain like any other part of a transaction, and screening and monitoring apply to the addresses, behaviour and exposure involved. What a fee can do is draw attention, which is why teams review it.
On 10 and 11 June 2020, two transfers of 0.55 ETH and 350 ETH each paid a fee of 10,668.73185 ETH, reportedly from the same address. SparkPool and Ethermine mined the two transactions. The reason for the fees was never conclusively established.
Neither pool announced a return. Ethermine’s operator announced on 15 June that it would distribute its fee to miners, and SparkPool announced on 16 June that it would do the same on 17 June. Whether SparkPool’s distribution took place was not confirmed.
The workflow has seven steps. Record the transaction and fee details, then compare the fee with the right baseline. Analyse the sender’s history and identify and screen the fee recipient. Trace the onward funds and check related wallets and exposure. Finally, document the findings and the escalation decision.
Every block now has a protocol-set base fee that is burned, plus an optional priority fee. The sender sets a maximum fee and is refunded anything above the base fee plus the tip. Since the Merge, validators rather than miners propose blocks, and block building is often handled by specialised builders. Since March 2024, blob transactions pay a separate blob fee.
The base fee is burned. The priority fee goes to the block’s fee recipient, which can be a validator’s address or a block builder’s address. Users can also pay the recipient through direct transfers inside a transaction, so a review should check both.
The 2020 incident is memorable for its size, but its lesson is practical. A fee that makes no economic sense is a prompt to ask who paid, who received and where the funds went next. On today’s Ethereum that also means asking which baseline applies, who the recipient really is and whether the payment came as a tip or a direct transfer. A fee anomaly is not an accusation, and it is not on its own a finding of financial crime. Teams that document their reasoning and apply a risk-based approach can handle these cases without overreacting or ignoring them.
To see wallet screening and transaction monitoring applied to your own flows, book a demo with the Scorechain team.















































