A practical walkthrough — how two complementary BIPs combine to enable permissionless, miner-secured sidechains on Bitcoin, using BTC as the asset, without federations or new tokens.
Drivechains = BIP-300 (Hashrate Escrows) + BIP-301 (Blind Merged Mining).
BIP-300 defines how value moves between L1 (mainchain) and L2 (sidechain). The design is deliberately asymmetric: deposits are cheap and instant to initiate; withdrawals are expensive to game and easy to audit.
You send BTC to a special OP_DRIVECHAIN output — the escrow UTXO tied to a specific sidechain slot. Equivalent coins then appear on the sidechain. From the user's point of view: it's a single Bitcoin transaction to a well-known address, and the L2 balance shows up on the other side.
Withdrawals don't happen one at a time. They're batched into bundles. Miners then vote (ACK) on each bundle over a long window — roughly 13,150 ACKs across ~26,300 blocks (about 3–6 months) are needed for a bundle to succeed.
The long window is the security feature, not a bug. It gives honest users months to notice a fraudulent bundle and coordinate against it. Miners who try to steal have to keep signalling for their theft in the open, block after block, for half a year — which makes the attempt trivially visible and its cost impossible to hide.
Sidechains occupy numbered slots (0–255). Anyone can propose a new one. Activation, like withdrawal bundles, involves miner ACKs. The slot number becomes the sidechain's stable identifier — Thunder, for example, lives at slot S9.
BIP-301 solves a different problem than BIP-300. Once the peg exists, who produces sidechain blocks, and how do they inherit Bitcoin's security? The answer is Blind Merged Mining (BMM).
Two roles are involved:
h*, the sidechain number, a reference to the previous mainchain block, and a fee offered to L1 miners.The result: exactly one sidechain block per mainchain block per sidechain, and sidechain fees flow back to Bitcoin miners as a fresh revenue stream.
Because miners are blind, sidechain operators have to compete on fee. The natural equilibrium is a small auction on every L1 block: whichever sidechain block producer offers the highest fee wins the right to have their block endorsed.
Thunder is a large-block payments/scaling sidechain, typically running at slot S9. It's essentially Bitcoin-style consensus rules but with much larger (and growing) blocks — starting around 10 MB and designed to scale toward hundreds of MB or 1 GB over time.
The point is high throughput for everyday payments while inheriting Bitcoin's hashrate security via BMM. It's one of the reference sidechains built by Layer Two Labs, alongside zSide (privacy), BitAssets, BitNames, and others.
There are two intervals to keep straight:
You can design a sidechain with a different internal target, but the canonical, secured blocks are still capped at one per L1 block via BMM. The official FAQ notes that changing the block time is possible but not recommended. Faster user-facing confirmation is better handled by mechanisms similar to Lightning-style channels or double-spend forfeits on top of already-confirmed sidechain state.
In practice — including on testnets and signets — Thunder block height tracks closely with the mainchain, consistent with the 1:1 BMM relationship.
No. This is one of the most useful things to internalize about Drivechains.
In shorthand:
This separation is the whole point of Blind Merged Mining. Sidechain activity stays overwhelmingly off L1 — only two things ever touch the mainchain:
Everything else — individual transactions, mempool churn, block validation — happens on the sidechain and never bloats L1. Bitcoin's throughput and node cost stay untouched, while high-throughput use cases get their own dedicated lanes secured by the same miners.
That's the trade Drivechains make: L1 gives up nothing, L2 gets Bitcoin's security in exchange for a tiny commitment per block and a slow-but-transparent withdrawal path. No federations. No new tokens. Just BTC on both sides, with miner-enforced rules doing the work of trust.