Carbon attribute and RE100 layer
Generated: 2026-05-08 (v3 BTM-numerator updated). Stage: post-Stage-4 commercial-case extension. Inputs: Ember Malaysia monthly CO2 intensity (measured 631.2 gCO2/kWh in 2024), 60 MW PV @ 100 MW DC reference scenario, 4 carbon price benchmarks.
carbon attributes are the dominant economic driver, not BTM For a 60 MW PV system at a 100 MW JB DC, displacing 46,647 tCO2/yr of grid emissions:
| Scenario | RM/tCO2 | Annual RM | 20y NPV (RM mn) | vs BTM-only NPV (+82.7 M) |
|---|---|---|---|---|
| BCX voluntary REC | 35 | 1.6 M | +16.0 | +19% |
| SG carbon tax 2024 (SGD 25) | 85 | 4.0 M | +38.9 | 47% |
| SG carbon tax 2030 path (SGD 65) | 230 | 10.7 M | +105.3 | 1.27× |
| Hyperscaler internal (USD 100) | 460 | 21.5 M | +210.7 | 2.5× |
| EU CBAM proxy (EUR 80–100) | 460 | 21.5 M | +210.7 | 2.5× |
At hyperscaler internal carbon prices (Microsoft USD 100/tCO2; Google ~80; AWS ~50), the carbon-attribute NPV alone exceeds the BTM avoided-cost NPV by 2.5×. If a hyperscaler PPA prices in carbon attributes — which they do, in 24/7 carbon-free energy commitments — carbon is the headline economic driver for the BTM PV case, not electricity savings.
Why this reframes the commercial story
Section titled “Why this reframes the commercial story”The BTM economics report tells the financial-risk story: PV alone banks 82.7 M NPV at 13.0% IRR (post-v3 partial-year fix). Solid, defensible, ICPT-driven.
The carbon analysis tells the strategic procurement story: the buyer (hyperscaler tenant) values the same kWh of PV at 0.2× to 2.5× its grid-displacement value because it advances RE100 / 24/7-CFE commitments that grid procurement at 631 gCO2/kWh cannot meet.
The two stories add — they are not substitutes. The total commercial value of 60 MW PV @ this DC under hyperscaler-internal-carbon-price contracting is:
BTM avoided cost NPV +82.7 M RMCarbon attribute NPV +210.7 M RM───────────────────────── ─────────Total stacked NPV +293.4 M RM (with ~13% pure-BTM IRR + carbon kicker)Implication: the floor on PPA-premium negotiation is not “avoided cost” — it’s “avoided cost + carbon attributes priced at tenant’s internal rate”. A hyperscaler that pays more than ~RM 350/MWh PPA premium is still cheaper than its internal carbon cost of buying fossil grid + offsets.
Carbon intensity context
Section titled “Carbon intensity context”| Country / region | gCO2/kWh | Source |
|---|---|---|
| Malaysia (2024 Ember) | 631 | this analysis, measured |
| Australia NEM (2024) | 600 | OpenNEM |
| Singapore | 400 | EMA |
| Global average | 430 | IEA |
| Germany | 350 | Fraunhofer |
| US grid average | 380 | EPA eGRID |
| Norway | 20 | NVE (hydro-dominant) |
Malaysia is moderately high-intensity — 50% above Singapore, 47% above the global average — driven by 44% coal + 34% gas. This makes each MWh of BTM PV displacement structurally more carbon-valuable in MY than in cleaner grids.
RE100 decomposition for the reference DC
Section titled “RE100 decomposition for the reference DC”A 100 MW DC at 80% utilisation × 1.4 PUE consumes ~981 GWh/yr. 60 MW BTM PV produces ~74 GWh/yr → covers 7.5% of consumption.
Going RE100 requires either:
| Path | Annual cost | NPV cost @ 8% × 20y |
|---|---|---|
| BTM PV (60 MW) for 7.5% + RECs for 92.5% gap @ RM 40/MWh | RM 36.3 M | −356 M |
| BTM PV (90 MW) for 11% + RECs for 89% @ RM 40 | ~RM 35 M | ~−344 M |
| 100% off-site PPA (CGPP/CRESS) at peninsular average ~RM 30/MWh premium | RM 29.4 M | ~−289 M |
| Stay on grid + Scope 2 offsets at hyperscaler internal RM 460/tCO2 | RM 285 M | ~−2,800 M |
Read: a hyperscaler paying its own internal carbon price for 100 MW of dirty Malaysian grid Scope 2 emissions is on the hook for 2.8 B RM over 20 years in offset-equivalent cost. Procuring renewable matching via a combination of on-site PV + off-site PPA at ~RM 290 M (8 of 100 of the offset cost) is the rational path.
The PV+REC stack also has a carbon-attribute auditability bonus — direct on-site generation is more defensible under SBTi / GHG Protocol Scope 2 location-based accounting than mere REC purchases.
What this means for the JS-SEZ commercial case
Section titled “What this means for the JS-SEZ commercial case”- The hyperscaler will pay for carbon, not for kWh savings. The commercial team’s pitch should lead with avoided emissions and RE100-fulfilment value, not with the (smaller) BTM NPV.
- PPA premium negotiation floor is ~RM 350/MWh, not zero. Any hyperscaler with an internal carbon price ≥ USD 80/tCO2 is paying more in offsets without the PV than they would with a moderate PPA premium for it.
- PV size sweet spot rises in the carbon framing. Per-MW carbon value is constant (~RM 359 k/MW/yr at USD 100/tCO2), so adding more PV adds proportionally to carbon NPV. The constraint is land / rooftop availability, not economics. The optimal PV size is the maximum the site can fit, not the BTM-NPV-optimised amount.
- BESS still doesn’t clear in carbon framing alone. BESS doesn’t add new clean kWh — it just shifts existing PV kWh in time. The per-MWh carbon attribute is the SAME with or without BESS. Carbon doesn’t fix BESS economics; it amplifies PV economics.
- 24/7 CFE is the upcoming standard, not flat annual matching. For 24/7 alignment, hyperscalers DO need BESS to time-shift PV into night-time demand. 24/7 CFE is the future BESS revenue stream that today’s BTM economics can’t price (no public market yet).
Sensitivity to grid intensity decarbonisation
Section titled “Sensitivity to grid intensity decarbonisation”If MY decarbonises faster than baseline (NETR target: 70% RE by 2050), grid intensity falls, and carbon value of BTM PV falls with it:
| Year | Projected MY grid gCO2/kWh | Annual avoided tCO2 | Annual value @ USD 100/tCO2 | NPV |
|---|---|---|---|---|
| 2024 (today) | 631 | 46,650 | 21.5 M | 211 M |
| 2030 NETR target | 500 | 36,950 | 17.0 M | 167 M |
| 2040 NETR target | 350 | 25,860 | 11.9 M | 117 M |
| 2050 NETR target (70% RE) | 200 | 14,780 | 6.8 M | 67 M |
The carbon NPV halves over 20 years if MY decarbonises on plan. This is a structural risk to “carbon-led” PPA pricing — but hyperscalers know this and price the future trajectory in their internal carbon prices today.
For commercial structures: lock in the high-carbon-value period (2025–2035) with a long-tenor PPA. By 2040+, MY grid is greener and PV’s marginal carbon value is materially lower.
Three commercial structures to consider
Section titled “Three commercial structures to consider”Structure 1: Pure BTM PPA (no carbon)
Section titled “Structure 1: Pure BTM PPA (no carbon)”- Buyer pays for kWh delivered at TNB-equivalent rate
- Seller absorbs carbon attributes
- Floor revenue: RM 270/MWh (avoided cost only)
- Suitable for: industrial customers with no carbon pressure
Structure 2: BTM PPA + Bundled RECs
Section titled “Structure 2: BTM PPA + Bundled RECs”- Buyer pays for kWh + receives carbon certificates
- Floor revenue: RM 310–350/MWh (avoided cost + REC)
- Suitable for: customers wanting standard 24/7 CFE annual matching
Structure 3: 24/7 CFE Premium PPA (recommended for hyperscalers)
Section titled “Structure 3: 24/7 CFE Premium PPA (recommended for hyperscalers)”- Buyer pays for kWh + carbon attributes priced at internal rate
- Includes BESS time-shift to maximise hourly carbon-free matching
- Floor revenue: RM 450–550/MWh (avoided cost + internal carbon + 24/7 premium)
- Reuses BESS as the “24/7 alignment” mechanism that BTM economics alone could not justify
- Suitable for: anchor hyperscaler tenants in JS-SEZ. This structure closes the BESS NPV gap by stacking the 24/7 premium on top of the carbon attribute value.
Structure 3’s 24/7 CFE premium ~RM 100/MWh at typical hyperscaler contracts. For 60 MW PV × 74 GWh/yr × RM 100/MWh = RM 7.4 M/yr. Modest, but stacked with carbon attribute monetisation closes the BESS gap when capex falls below ~3 M RM/MWh (i.e., by 2027–2028).
What’s still uncertain
Section titled “What’s still uncertain”- Hyperscaler internal prices vary. Microsoft USD 100/tCO2 is the upper bound; AWS USD 50 is the lower. PPA pricing depends on which tenant. Use USD 75/tCO2 as a middle-of-the-market planning value.
- 24/7 CFE is still nascent. Google’s first 24/7 CFE deals are ~2022. Pricing precedent is thin in MY. Negotiate a 24/7 premium separately from base PPA so it’s auditable and re-negotiable.
- REC market in MY is shallow. BCX is voluntary; mandatory RPS does not exist. RECs may be discounted or non-monetisable.
- Scope 2 accounting standards are tightening. SBTi 24/7 CFE guidance is in development; GHG Protocol Scope 2 amendment was published 2024-Q4. Long PPAs may need amendment provisions.
- This report:
reports/carbon_re100_analysis.md - Raw evaluation:
reports/carbon_re100.json - Source code:
src/jb_vpp/models/carbon.py - Cross-references:
reports/btm_economics_dc100.md,reports/vpp_service_revenue_required.md,reports/sensitivity_tornado.md,reports/aggregator_portfolio.md
Caveats
Section titled “Caveats”- Linear monetisation: assumes carbon attributes are saleable at the scenario rate every year for 20 years. In practice rates evolve and carbon-attribute markets fluctuate.
- Static MY grid intensity (does not project NETR path within the 20-year NPV horizon — see “sensitivity to decarbonisation” above for the impact, which is not folded into the headline NPV).
- 24/7 CFE premium of RM 100/MWh is a planning estimate, not a market-cleared price.
- All carbon NPV figures stack ON TOP of BTM avoided-cost NPV; do not double-count.