Representative public model · South Africa · Corporate power

A corporate PPA is not cheap because the headline energy price is low. It is cheap only when the buyer’s full delivered bill is lower after every network and settlement layer.

Felix built a representative 50 MWac generator-to-corporate-buyer model to separate the PPA price from wheeling losses, use-of-system charges, balancing, residual grid supply and profile mismatch. It is a public analytical model, not a completed client mandate or municipal quotation.

50 MWac generator120 GWh Year 1 buyer load15-year representative PPABuyer and generator views separated

The modelling problem

The same transaction can look attractive to the generator and poor to the buyer

A generator sees contracted volume multiplied by the PPA tariff. A corporate buyer sees the PPA invoice plus network charges, losses, service fees, residual grid energy, demand charges and settlement adjustments. A trader or municipality may see another layer of balancing, credit and account administration.

The model was therefore designed as two reconciled cases rather than one blended project return. Buyer savings could not be used to justify generator IRR, and generator revenue could not be treated as the buyer’s all-in cost.

Analyst note

The most common commercial error is to compare a PPA tariff with a grid tariff. Those are not like-for-like numbers. The comparison becomes meaningful only after the wheeled energy reaches the buyer’s account and the residual bill is rebuilt.

What a live mandate would need before any saving is quoted

Half-hourly load and generation profiles, not only annual MWh.
The applicable Eskom or municipal tariff schedule and demand charges.
Connection point, wheeling path, loss factor and use-of-system charges.
Trader, balancing, billing and settlement fees.
Residual-supply terms and treatment of surplus or shortfall energy.
PPA escalation, credit support, termination and change-in-law provisions.

The cost bridge

The headline tariff was less than four-fifths of delivered renewable cost

Cost component	Representative treatment	Why it matters
PPA energy	R1.15/kWh	The generator’s contracted price, not the buyer’s total cost.
Network and use-of-system charges	Added separately	Can vary by injection point, customer account, municipality and tariff year.
Losses	Reduce delivered MWh	The buyer may pay for generated energy but receive less at the meter.
Balancing and services	Separate fee layer	Covers forecasting, nominations, imbalance and settlement functions.
Residual grid supply	24 GWh in Year 1	The buyer remains exposed to grid energy and demand charges when profiles do not match.
Surplus / shortfall	Requires explicit settlement rule	Annual renewable share hides the value and cost of energy at the wrong time.

Human judgement

A 17.3% saving is not a market claim. It is the output of one explicit set of assumptions. The useful result is knowing which tariff, profile and network inputs can eliminate that saving before the parties sign.

Year 1 bill reconciliation

Buyer view

Grid-only baseline: R222.0m
PPA energy: R110.4m
Network and services: R28.8m
Residual grid supply: R44.4m
Total with wheeling: R183.6m
Illustrative saving: R38.4m

What breaks the buyer case

Savings were most fragile where the project team had the least control

Felix tested the commercial logic across PPA price, avoided grid tariff, delivered energy and network charges. The model made it possible to see whether value was genuinely shared or whether one party’s downside was simply being transferred to another.

Stress	Commercial effect	Negotiation implication
PPA tariff +10%	Buyer saving falls while generator return improves.	Agree a value-sharing range rather than negotiate the energy tariff in isolation.
Grid tariff -10%	The avoided-cost benchmark weakens.	Do not rely on perpetual grid escalation without a transparent tariff path.
Delivered energy -10%	Generator economics weaken and residual supply increases.	Profile and availability risk need separate treatment from annual yield.
Network charges +20%	Buyer saving is compressed without improving generator revenue.	Pass-through, reopeners and savings floors become material.
Combined downside	Illustrative buyer saving approaches 2%.	Transaction complexity may no longer justify implementation unless other value exists.

The contract questions

Every model line needed an owner in the term sheet

Volume and profile

Who owns mismatch?

Annual generation can exceed annual demand while the buyer still needs significant residual power. The PPA must deal with surplus energy, shortfall, balancing and portfolio allocation.

Evidence requiredInterval profiles and settlement waterfall

Grid and curtailment

Who pays when the network cannot deliver?

Deemed energy, network outage, curtailment caps, force majeure and relief events determine whether lost generation becomes a generator, buyer or network risk.

Evidence requiredConnection and wheeling agreements

Buyer credit

What survives an early termination?

The generator needs payment security, termination compensation and replacement-buyer rights that match the debt profile and break costs.

Evidence requiredGuarantee, LC and termination schedule

Metering and settlement

Which meter controls the invoice?

Meter hierarchy, data timing, adjustments, disputes and audit rights determine when calculated savings become cash.

Evidence requiredMetering protocol and bill examples

Disclosure boundary

This is an independent representative model. The tariffs, charges, capacity and savings are synthetic and do not represent a quotation, live client transaction, available network capacity or legal assessment. A live case must use the current provider, municipal, trader and contract schedules.

Corporate PPA analysis

Need the delivered-cost bridge built from actual bills and interval data?

Share the buyer accounts, tariff schedules, load data, generator case and draft commercial structure. The output should show who creates value, who carries risk and where the saving disappears.

Request a scope Inspect the public model