Tron (TRX) sustainability report

TRON is present on the following networks: Tron.

The TRON network reaches agreement through delegated proof of stake. Holders of the native asset lock it up, which yields voting weight, and use that weight to back candidates who have registered to produce blocks. Votes are counted afresh at the end of every six-hour cycle. The twenty-seven candidates ranked highest by votes become the super representatives that produce blocks for the cycle that follows; those ranked immediately below them form a standby tier of partners, which does not produce blocks but remains in the reward distribution and supplies replacements as the ranking shifts. Because the count repeats four times a day, the producing set is re-derived continuously rather than fixed for a term.

Block production follows a fixed rotation among the twenty-seven, with one slot every three seconds. A producer that misses its slot is skipped and the schedule continues. A block is treated as settled once more than two-thirds of the producing set has built on it, so settlement follows from producers confirming one another's work rather than from a separate voting protocol, and takes on the order of a minute in normal conditions. Smart contracts execute in a virtual machine built for compatibility with Ethereum tooling, under the same producing set.

The security argument rests on an honest supermajority within a small, elected and publicly identified group. Entry is open in the sense that anyone may register as a candidate, subject to a deposit that is destroyed on registration, but a candidate only produces blocks by accumulating votes. The same twenty-seven form the committee that governs the chain's adjustable parameters: proposals to change values such as resource pricing, reward sizes and protocol feature switches are voted on by the producers, and a proposal passes on the support of a supermajority of them. A substantial part of what users experience as the cost of using the network is therefore a governance variable rather than a fixed property of the protocol.

TRON is present on the following networks: Tron.

Producers on TRON are paid from two streams of newly issued units. A fixed amount is awarded for each block to the representative that produced it, and a larger per-block amount is divided among the wider set of elected and standby representatives in proportion to the votes each received. Each representative publishes the proportion of its receipts that it passes back to those who voted for it, so what a participant earns depends on which representative is backed. Registering as a candidate requires a deposit that is destroyed rather than held, which discourages frivolous entry. There is no slashing: a representative that produces poorly is not deprived of locked funds but loses votes, and with them its place in the producing set and its share of both streams.

Users do not pay a per-transaction price in the usual sense. The protocol meters two resources. Bandwidth covers the byte size of a transaction and energy covers computation performed by the virtual machine. Each has a fixed network-wide daily ceiling, and locking up the native asset entitles an account to a share of that ceiling proportional to its share of everything locked for the same resource, so an allowance changes as others lock and unlock even when the account itself does nothing. Every account also receives a small free bandwidth allowance that regenerates daily. Under the staking arrangement introduced in 2023 and generally known as the second version, locking up is separated from resource assignment: resources obtained by locking can be delegated to other addresses and reclaimed without unlocking, which is what makes third-party resource provision practical. Recovering the locked balance itself requires a waiting period of fourteen days.

When an account attempts an operation without sufficient resources, the protocol burns the native asset at unit prices set by governance, one per byte of bandwidth and one per unit of energy, both of which have been revised upward in recent parameter changes. That burn is destroyed rather than paid to producers, so what users spend and what producers earn are separate flows. A further mechanism adjusts cost by contract rather than by congestion: a contract whose energy consumption passes a threshold within a cycle carries a multiplier on its energy cost in following cycles, decaying once consumption falls back, so calling a heavily used contract can cost several times what the same computation costs elsewhere.

TRON is present on the following networks: Tron.

The consumption figure reported for this network is estimated from the machines that operate it. A delegated proof-of-stake chain does not expend energy as part of reaching agreement, so there is no work-based quantity to model as there is for mining networks; what consumes electricity is a population of continuously running servers, and estimating that population is the whole of the exercise.

The estimation approach used here starts from the participants the chain itself identifies. Registered producer candidates are visible on chain, as is their ranking, which separates the twenty-seven producing during a cycle from the standby tier and the wider candidate list. Around that core sits a larger population of full nodes, relay nodes and the query-serving infrastructure that applications and wallets depend on, which is estimated from publicly available network data and from scanning for reachable endpoints. A representative machine is then inferred for each part of the population, taking the published operating requirements of the node software as the primary input. Those requirements are demanding relative to many networks, reflecting a three-second block interval, a high sustained transaction rate and a large accumulated state that nodes must keep available. Electrical draw for machines of that class is taken from controlled bench measurement rather than from specification sheets, and includes the draw of a machine that is powered and connected but idle, which for always-on infrastructure is a large part of the annual figure. The total is the sum across the estimated population.

Where a figure is needed for one asset issued on the chain rather than for the chain as a whole, a share of the network total is attributed to it in proportion to observed on-chain activity involving that asset. This matters on a network carrying a high volume of token transfers relative to its other traffic.

The limits are inherent to the method. Node counts and hardware profiles are inferred from public observation and stated requirements, not metered; operators commonly run hardware above the published minimum; and endpoints that do not respond to scanning are not counted. Where evidence is thin, assumptions are chosen to be more likely to overstate impact than to understate it, and figures are revised as observation improves.

TRON is present on the following networks: Tron.

The renewable proportion reported for this network follows from where its machines run. Locations are inferred from publicly observable network data: the addresses of reachable nodes are resolved to hosting providers and regions, and the on-chain registry of producer candidates helps place the most important part of the population, since candidates campaign for votes publicly and many disclose who operates them and where. The producing set is small and identifiable, which makes the part of the network that matters most for block production easier to locate than the broader population of supporting and query-serving nodes.

Where the geographic spread of that broader population cannot be established from observation, the distribution of a network with a comparable validator selection and reward design is used as a substitute. This substitution is the main source of uncertainty in the reported share and should be read as such.

Each location is then matched to published statistics on how electricity is generated on the grid serving it, and the individual shares are weighted by the electricity attributed to the machines in that location to give a network-level proportion. The underlying statistics describe the average generation mix on a regional grid across a reporting year. They do not capture variation within a day or a season, nor any renewable supply contracted privately by an individual operator, which is not observable from the chain.

Energy intensity is stated as a marginal figure: the additional electricity associated with processing one more transaction, rather than annual consumption divided by transaction count. On a network whose servers run continuously and which sustains a high transaction volume, that marginal quantity is very small, and it is sensitive to the throughput assumed in deriving it.

The generation statistics are taken from Share of electricity generated by renewables, compiled by Ember and the Energy Institute's Statistical Review of World Energy and processed by Our World in Data.

TRON is present on the following networks: Tron.

Emissions attributed to this network are derived from the geographic picture built for its energy mix, applied to a different statistic. Once the machines have been placed in regions, the electricity attributed to each region is multiplied by the carbon intensity of generation on the grid serving it, expressed as emissions per unit of electricity delivered, and the results are summed across regions. Where node geography cannot be observed directly, the distribution of a structurally comparable network stands in, and that assumption propagates into the emissions figure exactly as it does into the renewable share.

The figures separate two categories. Scope 1 covers emissions from sources the operators of the network's infrastructure control directly, such as fuel burned on site; for servers hosted in commercial facilities drawing from public grids, this is ordinarily nil or close to it, with any backup generation contributing negligibly over a reporting year. Scope 2 covers the indirect emissions embodied in the purchased electricity that powers those servers, and represents effectively the whole footprint of a network of this design. The allocation is made from grid electricity drawn and is not adjusted for offsets, attribute certificates or supply agreements held by individual operators, since none of those are visible from the chain.

Greenhouse gas intensity is reported on a marginal basis, as the emissions associated with one additional transaction rather than an average obtained by dividing an annual total. Because the infrastructure draws power whether or not transactions arrive, this marginal quantity is small, and it moves with the throughput assumed at least as much as with anything the protocol does differently from one period to the next.

Carbon intensity values are taken from Carbon intensity of electricity generation, compiled by Ember and the Energy Institute's Statistical Review of World Energy with major processing by Our World in Data, and made available under the CC BY 4.0 license.