Investment
Pre-connection models keep moving
A change to tariff, utilisation or capex forces a rebuild, and versions scatter across files.
- Assumptions unversioned
- Recalculation by hand
Sector · Renewables & infrastructure
Returns cannot be reviewed while the model and the O&M ledger stay apart
Disputes over renewable assets concentrate in one place: the feasibility model assumes one tariff and one utilisation figure, operations record another set of generation and downtime. Keep the two apart and a shortfall can never be attributed. We connect feasibility, construction and O&M on a single ledger so assumptions and actuals sit on the same timeline.
Context and source of the rules
Definitions come from tariff, subsidy and availability rules
For renewables and infrastructure the key metrics are externally set: tariff and subsidy rules, grid-connection and dispatch definitions, availability metrics. Those three decide how revenue is computed, how availability is counted, and at what point a project is deemed to have reached designed output.
Those definitions also move. Tariff policy and subsidy phase-down adjust annually; utilisation hours vary with resource conditions; availability is constrained by recorded downtime. The assumptions frozen at feasibility part company with recorded reality by the third operating year, and most ledgers still hold only the original model.
Our position: model feasibility and operations on the same objects. Project, unit, grid-connection point, tariff, generation and downtime share one identifier; assumptions keep their versions, operating data keeps its source, and post-investment review then has a comparable basis.
One source of truth
A model table, an operations table, one project key
The usual failure shape: the financial model lives in a spreadsheet, the O&M ledger lives in the production system, and the two meet only when someone assembles the annual report. A gap assembled by hand cannot be attributed — you cannot tell whether the tariff moved, the resource fell short, or equipment was down.
We decompose the model into versioned assumption records (tariff, utilisation hours, capital expenditure, operating cost), collect operations data against the same master key, and align the two by period in review. Variance then splits into price, resource and operations rather than one line reading "below plan".
- Assumptions stored as versions, changes traceable
- Generation, outage and maintenance tied to project
- Tariff and subsidy handled as fields, not footnotes
- Review outputs an attributable variance breakdown
Live pressure
Four roles demand the same number
Definition conflicts in renewables surface together at grid connection, audit and refinancing.
Operations
Generation detached from book revenue
Generation and outage sit on the production side, revenue on the finance side; they do not reconcile.
- Two record sets
- Variance unattributable
Audit
Designed-output tests unproven
Reaching design output is asserted from manual roll-ups without continuous same-basis records.
- Broken definitions
- Process not reconstructable
Asset
Sites logged at different depths
Central plants, distributed projects and infrastructure assets keep ledgers of different granularity.
- Uneven site basis
- Roll-up needs mapping
The point of renewable asset software is not more telemetry. It is putting telemetry and the original investment assumptions inside one set of field definitions.Delivery methodology · indicator framework
Order of remediation
Start with definitions, close the loop at review
Define metrics, then release workflow, then connect operating data. Reversed, the operating history lands before the basis is fixed and has to be cleaned twice.
01 · basis
Version the metrics
Fix how tariff, utilisation hours, availability and designed output are computed, and who owns each.
- Definition table
- Field owners
- Period basis
02 · model
Versioned feasibility
Treat the model as an object: assumptions, results and the approving decision kept as one version.
- Assumption history
- Sensitivity trail
- Linked to approval
03 · build
Milestones and cost
Progress, change orders and capital expenditure accrue against the same project key as the asset base.
- Milestone trail
- Capex mapping
- Capitalisation basis
04 · operate
Telemetry into the ledger
Generation, outage and maintenance records post to project and unit, aligned to model fields.
- One master key
- Sourced records
- Anomalies visible
05 · review
Attribute the variance
Compare model and actual by period and split the gap into price, resource and operations.
- Same-axis compare
- Variance split
- Feeds next model
Capability map: site question to system object
Mapped to the published capability labels and system catalogue. Fields and flows are configured to your definitions; this is not a promise about specific collection interfaces.
| What the site asks | System object | Capability label | Output |
|---|---|---|---|
| Which model version did this project rely on | Model object with assumption versions | Feasibility modelling | Versioned model record |
| Which projects does a tariff change touch | Tariff fields plus asset relationships | Infrastructure investment system | List of affected projects |
| Do progress and spend match | Milestones and cost accrual | Construction management | Progress-to-cost table |
| Why is this month’s generation below plan | Operating records and outage ledger | O&M monitoring | Price / resource / operations split |
| Has the project met its designed-output test | Availability metric with period calculation | O&M monitoring, feasibility | Evidence on one basis |
| Can all sites roll up into one export | Shared master data and aggregation rules | Infrastructure investment system | Group-level asset ledger |
Asset types and deployment forms
Asset types follow the product description on the company site. Collection and integration methods are confirmed per site during assessment.
SolarWindHydroStorageInfrastructure assetsCentral plantsDistributed projectsPrivate cloudAir-gapped intranetDingTalk / Lark / WeComFinancial-report OCRFunction / data / field permissions
State the evidence boundary first
A product segment, not a separately ranked line
Two claims on this page are substantiable: the renewables management platform appears among the four platforms on the company site, and the infrastructure investment management system appears in the published catalogue of business systems.
What we do not claim: the third-party rankings and public case references are not assessments of the renewables segment, so we do not use them as sector evidence here. Where an operator needs delivery references, we discuss them separately by site type and collection conditions during assessment.
10+
Years of industry experience (since 2016)
60%+
Share of staff in engineering
15
Domestic offices for on-site delivery
100+
Registered software copyrights
Substantiable basis and service terms
Outcome-style figures are stated only as cumulative platform coverage, so they are not read as assets we manage or returns we produced.
| Item | Basis | Qualifier |
|---|---|---|
| Product form | Renewables platform (one of four) / infrastructure investment system | Stated per the published product architecture and catalogue |
| Capability stages | Feasibility / construction / O&M | Site labels; depth set by project type |
| Platform scale | 14,000+ funds / RMB 5.6 tn+ covered | Managed-basis coverage, not proprietary assets or returns |
| Security | AES-256 at rest, TLS 1.3 in transit, three permission levels | Per deployment form and contract terms |
| Response | 24/7 response, issue located within one hour, one year free after acceptance | Contract terms |
| Certification | CMMI Level 3, ISO 9001, high-tech enterprise, Xinchuang ready | Scope per certificate |
When we advise against buying
A single-plant owner that already runs a usable monitoring platform usually needs a mapping exercise between operating data and finance definitions; a metric crosswalk solves that more cheaply. We also decline briefings where software is expected to settle tariff, subsidy or grid-connection questions. Those conclusions belong to policy research and counsel; the platform keeps the evidence and the result traceable.
Questions at the shortlist stage
Mechanism answers; the governing tariff, subsidy and connection positions stay with your jurisdiction.
How is a pre-connection model handled without rebuilding it every time an assumption moves?
Turn the model from a sheet into a versioned object. Tariff, utilisation hours, capital expenditure and operating cost become fields; each adjustment creates a new version holding the author and the reason. Results recompute per version and the approval links to one of them. When a tariff change lands, affected projects are queried by field and rerun, instead of every model being rebuilt by hand.
Actual generation falls well short of feasibility. Can the system locate the cause?
It can decompose the gap, provided both sides share a basis. Utilisation hours and book revenue in the feasibility case correspond to generation, availability and recorded outage on the operations side. Aligned by period, variance splits into price, resource and operations. The platform supplies the decomposition, not the verdict — whether the resource underperformed or downtime exceeded plan still needs a business judgement, but that judgement now rests on retrievable evidence.
When tariff or subsidy rules change, should historical reports be restated?
Not silently. Keep the finalised historical basis as its own version, add the policy effective date, and let reporting produce a then-basis and a current-basis figure with the difference labelled. Immutable history plus an effective-dated new basis is what both audit and investment review accept.
How uniform must ledgers be across several site types?
Unify three layers — the project master key, the metric definitions, and the period basis — and leave collection fields configurable per central plant, distributed project or infrastructure asset. Roll-ups align on the three shared layers while detail differences stay inside their own view. Forcing field-level uniformity typically makes lightweight assets carry heavyweight recording costs.
We already have a production monitoring system. Why add an investment platform?
Test it with one question: can you see, in one place, which assumptions the model was built on and what the asset is actually delivering? Production monitoring covers equipment state; it does not carry versioned investment assumptions or post-investment review. If that comparison is still assembled by hand, the two systems are complements.
Next step
Compare one feasibility model against one operating month
In a 45-minute session we check whether model assumptions and O&M metrics share an axis, name the three layers that need versioning, and propose the release order. Our insight notes cover the same definitional questions.