DC MOC — Methodology & Formulas

01 Purpose & engineering basis

DC MOC (Data-Centre Management & Ops Cost) is a PRO-tier Next.js static-export SPA that answers five operator questions in a single instrument surface: what will this facility cost to build (CAPEX), what will it cost to run per year (OPEX), how should operations be staffed across shifts, what is the probabilistic range of commissioning cost (Monte Carlo), and how is this summarised for executive sign-off (PDF export). Access is governed by the site-wide auth-tier matrix at the PRO level.

The mathematical basis is deterministic-by-default: every KPI is sourced from a locked constant or a formula anchored to a published standard, cost index, or RZEngine DATA entry. Monte Carlo randomness is seeded where stochastic output is needed and produces reproducible results. The tool is an engineering pre-design aid and Class 4 budgetary estimate (AACE 18R-97); detailed engineering procurement must follow.

The SPA consumes two shared engines loaded at page root as deferred scripts: /rz-engine.min.js (exposes window.RZEngine + window.RZEngine.DATA) and the DCMOC TypeScript bundle. During Next.js SSR prerender, dcmoc/src/lib/rz-engine.ts eval-requires the CommonJS engine from the repo root so that server-rendered and client values agree with zero hydration mismatch.

02 Inputs

Fourteen primary inputs span five groups — project identity, site/location, electrical system, cooling/mechanical, and staffing. Additional advanced fields (seismic zone, substation type, green certification, renewable option, deep-sea cooling parameters) unlock in PRO mode. All ranges are enforced client-side; missing optional fields fall back to engine defaults.

03 Calculation methodology

Five calculation layers run in sequence. Function names and file paths map one-to-one to the source noted in each formula block.

3.1 CAPEX build-up

3.2 PUE & timeline

3.3 OPEX — total annual cost

3.4 Shift & staffing optimisation

3.5 Commissioning program & Monte Carlo

3.6 General Monte Carlo driver (sim.monteCarlo)

04 Constants & data sources

All constants are deep-frozen in RZEngine.DATA (loaded from rz-engine.min.js) or in dcmoc/src/lib/capex-data.ts (DCMOC-finer A7 2025 bases). Key tables:

CAPEX base cost factors — A7 2025 rates

Location multipliers (fallback; priority = constructionIndex)

Staffing & OPEX constants

05 Outputs

06 Worked example

A 2 000 kW Tier III purpose-built data centre in Jakarta (DCMOC city market: jakarta, $/W anchor 11.21 T&T 2025), N+1 redundancy, in-row cooling, standard racks, modular UPS, diesel generator, 24 h fuel, 2026 build year, 30 FTE in-house staff, 60% PM ratio, 85% retention. Values computed from source constants — engine cannot be require()-called in this static doc, so the trace below follows the exact formula path from source.

07 References & standards

• T&T DCCI 2025 — Turner & Townsend Data Center Construction Cost Index 2025: per-city $/W anchor table (Singapore 14.53, Tokyo 15.20, Jakarta 11.21, London 12.00 …); country construction indices. Primary CAPEX cost anchor.
• C&W 2025 — Cushman & Wakefield DC Cost Guide 2025: US city $/W checks (Virginia 13.40, Dallas 14.30, San Antonio 9.30); supplementary to T&T.
• JLL 2026 — JLL Data Center Construction Cost 2025–2026: forward escalation and phase premium (~12% per phase). Year-escalation multipliers in engine.
• AACE 18R-97 — Cost Estimate Classification System: DC MOC CAPEX is a Class 4 budgetary estimate (−30/+50% accuracy range). Class 5 (-50/+100%) when fewer than 5 inputs are defined.
• ASHRAE TC9.9 (5th Ed., 2021) — Thermal Guidelines for Data Processing Environments: A1–A4 air-class and H1 liquid supply-temperature envelopes; ΔT limits driving PUE bands.
• Uptime Institute Tier Standard: Topology 2022 — T1–T4 redundancy path definition; Tier availability targets (99.671% … 99.995%); IST scenario counts driving Cx level-5 duration.
• ANSI/TIA-942-C (2024) — Rated-1 … Rated-4 DC infrastructure classifications; topology cross-reference for DCMOC architecture module.
• ASHRAE Guideline 0-2019 — The Commissioning Process: Cx level definitions (L0 pre-design … L6 post-occupancy) that map to the engine's Cx duration model.
• BCxA / NETA ECS — Building Commissioning Association and National Electrical Testing Association scope benchmarks; calibrate the Cx base day-rate cards and level cost ratios.
• ISA-18.2 / EEMUA-191 — Alarm management targets (≤1 alarm/10 min/operator; flood ≥10 alarms/10 min; actionable ≥85%) embedded in RZEngine.models.alarms.
• SHRM 2024 — Employee replacement cost 150% of annual salary; feeds retentionCost in the staffing resilience model.
• IEA 2026 + Ember — Grid carbon intensity by country in DATA.carbon.gridFactor; ICAP carbon prices in DATA.carbon.carbonPrice. Drive extended-OPEX carbon line.
• Engine source — rz-engine.js (RZEngine v2.5.1, DATA v2.5.1) · dcmoc/src/lib/CapexEngine.ts · dcmoc/src/lib/capex-data.ts · dcmoc/src/lib/engine-catalog.json (200 functions / 105 sources, AUTO-GENERATED).

08 Assumptions & limitations

CAPEX is a Class 4 budgetary estimate (AACE 18R-97) with ±30–50% accuracy. It is not a detailed cost plan and must not be used for contract tendering without a full bottom-up quantity take-off by a qualified cost estimator. Key assumptions:

CAPEX: The 14-category base costs are 2025 market rates calibrated to the T&T / C&W city-anchor tables. The DCMOC A7 2025 electrical base (1 550 $/kW) is deliberately NOT merged with the shared engine's 1 200 $/kW model — both are documented in engine-catalog provenance as a justified divergence (different model shapes). Seismic multiplier defaults to zone 2 (moderate seismicity) when not specified. Green certification and renewable options add premium costs not captured in the base categories — handled via renewableCost adder. FOM (Facilities of Merit — substation, grid connection, switchgear) is an optional line that requires project-specific utility quotes to confirm.

OPEX: DC-contract utilisation basis assumes 100% committed capacity at all times (not a measured blended utilisation). Energy price is the regional benchmark rate from DATA.regions[*].powerKwh — actual tariff structures (TOU, demand charges, PPA) can be overridden via advanced inputs. Maintenance requires a CAPEX basis; the model computes zero maintenance if no capex is provided.

Staffing / shift: The 4.2× shift multiplier is an Uptime-benchmarked industry norm for critical-facility 24/7 coverage. Actual headcount will vary with site complexity, automation level, and local labour regulations. The Staffing Resilience Index (SRI) is a screening heuristic — it is not a certified HR assessment tool.

Monte Carlo: The commissioning Monte Carlo perturbs only IT load (±7.5%) and a scalar cost noise (±5%). It does not model schedule risk, vendor variation, or supply-chain disruptions. For project-level risk analysis a full probabilistic schedule with correlated distributions is required. The sim.monteCarlo pairwise-correlation implementation is exact for two independent pairs; chained pairs introduce implicit transitive correlation — use disjoint pairs only.

PDF export: The executive PDF is generated via the browser print-window (ReportNarrative engine). Chart rasterisation fidelity depends on the browser renderer; vector PDF export is not supported in the current release. This manual and the DC MOC tool are an engineering education and pre-design aid; financial investment decisions must engage qualified engineers, cost consultants, and financial advisers.