SINGAPORE, 8 AUG 2026 — An MP asked how long a completed solar installation waits before it is connected to the grid. The Trade and Industry Ministry replied that it does not have the data, and explained why: nobody is required to report when an installation is finished.
The grid is not what is missing here. The measurement is.
The question and the answer
Ms Valerie Lee asked, on 4 August, what the average time has been over the past two years between the technical completion of solar photovoltaic systems and their grid connection, separately for residential and industrial installations, and what is being done to shorten that interval so generated electricity can be exported promptly.
Minister for Trade and Industry (Energy and Industry) Dr Tan See Leng answered the first part in a single sentence. The government does not have data on the lag time between completion and grid connection, because solar installers are not required to report the actual date of completion to the government or to SP Group.
The rest of the reply describes the connection process, not how long it takes.
What is measurable, and from when
Every clock the reply describes starts at the application.
| Installation | Process after completion | Stated timing |
|---|---|---|
| Embedded solar PV below 1 MWac | A Licensed Electrical Worker certifies the installation and should promptly notify SP Group | SP Group can inspect and turn on within 10 calendar days of receiving a complete application |
| 1 MWac and above, or High Tension connected | The LEW submits a turn-on application nominating appointment dates; inspection and turn-on happen at the appointment; a completed declaration follows | Preferred dates at least 4 business days away; declaration within 7 business days |
From the Ministry of Trade and Industry's written reply of 4 August 2026. These are the timings the reply states for the steps after a complete application is received. They are not measurements of how long installations actually wait, which the reply says is not recorded.
Read carefully, the ten-day figure is conditional twice over. It applies once SP Group has received a complete application, and it is what SP Group can do rather than what is observed. For larger installations an SP PowerGrid officer may need to attend commissioning, and additional technical and regulatory requirements apply.
These are reasonable requirements, and they measure the application process. An owner's wait starts earlier, when the panels are finished and the system is sitting there doing nothing.
Why the missing data is the finding
The reply is candid about which part of the sequence nobody records, and its remedies follow from that candour.
The government's remedies are addressed to behaviour rather than to performance: it encourages timely commencement of the grid connection process, prompt completion of applications and technical requirements, engagement with SP Group on feasibility before installation begins, and prompt completion of remaining procedures afterwards. It will continue working with SP Group and industry stakeholders to explore ways to expedite connection.
Encouragement is what is available when the delay cannot be located. If the lag sits between completion and the installer filing a complete application, no service standard on SP Group touches it, and no data would reveal it.
The structural point is that the one participant with no reporting obligation — the installer — is also the one who starts the unmeasured interval.
What idle capacity costs
While the delay is a minor cost for any single system, the aggregate cost is significant.
A completed rooftop array that is not yet connected generates nothing. The owner has paid for the equipment and the installation, and the payback period runs from the day the meter turns, not the day the panels go up. For a household the effect is an irritation; for an industrial site with a large array it is capital sitting idle.
There is also a system-level effect. Singapore's solar targets are expressed in installed capacity, and capacity that is installed but unconnected counts toward the target while contributing nothing to generation. Those two quantities diverge by exactly the amount nobody records.
Nothing in the reply says the lag is long, and it may well be short in most cases. The honest position is that neither we nor the ministry can say.
The fix is cheap and specific
A reporting requirement on the completion date would close the gap, and it is a small ask.
The Licensed Electrical Worker already certifies the installation and already notifies SP Group. Recording the completion date at that point would create the missing series without adding a step, and it would let the ministry answer next year the question it could not answer this year.
Whether that is worth doing depends on how large the lag turns out to be, which is the circularity the current arrangement produces: the case for measuring cannot be made without the measurement.
Who each clock belongs to
The connection sequence involves three participants, and accountability follows the handoffs between them.
The installer builds the system. The Licensed Electrical Worker tests, certifies and notifies. SP Group inspects and energises. The stated service standards attach to the last of those three, and the reporting obligations attach to none of the first.
An owner experiencing a long wait therefore has no way to establish where it arose. The reply's advice — engage SP Group on feasibility before installation begins — is a reasonable hedge against one specific failure, which is discovering after the fact that connection is not straightforward at that site.
It also asks the owner to manage a process whose timings they cannot see.
Why this is a common shape in energy policy
This pattern of unmeasured handoffs generalises beyond solar installations.
Where a policy outcome depends on a chain of private actors and the state regulates only the final step, the state's data covers only the final step. Service standards get published for the regulated interface and the rest of the chain is invisible.
Deployment targets are then set in units the state can count — installed capacity, applications approved — rather than in the unit that matters, which is electricity actually delivered. The two diverge exactly in the gap nobody measures.
None of this requires anybody to be at fault. It is what happens when the reporting boundary and the process boundary are drawn in different places.
The question the reply answers well
To its credit, the reply is transparent about what it does not know.
A ministry that does not hold data has two options: say so, or produce a number from an adjacent series and let the reader assume it answers the question. The second is common and the first is better, and this reply took the first.
It also explains precisely why the data is absent — no reporting obligation on installers — turning an unanswerable question into a specific, identifiable gap. That is more useful than an estimate would have been, because it points at the change that would fix it.
The issue is not that the ministry dodged the question, but that this data gap has presumably existed for years and was only surfaced by a parliamentary inquiry.
What to watch
Two things.
The first is whether a reporting requirement appears. The reply commits to exploring ways to expedite the process; a completion-date field would be the smallest concrete version of that.
The second is whether the question returns. Parliamentary questions that produce "we do not have the data" are frequently asked again, and the second asking is usually harder to answer the same way.
View author profile → · Editorial policy
