How KAPSARC government rates flow into your BOQ

KAPSARC — the King Abdullah Petroleum Studies and Research Center — publishes more useful data for construction estimators than most estimators realise. Energy unit costs, transport fuel benchmarks, material price indices, regional cost factors. The data is open, structured and reasonably current.

It also tends to be ignored, because it sits behind a portal that nobody wants to navigate at tender time. ORKSTRA pulls it in automatically. This post explains what KAPSARC covers, how it is wired into the BOQ pricing engine, and what to do when you are working outside its scope.

What KAPSARC publishes that matters to a QS

KAPSARC's open dataset is wider than the energy focus implies. The parts that touch construction estimating include:

• Energy unit costs. Electricity, diesel, gas — by region and by month.
• Material cost indices. Cement, steel rebar, aggregate — KSA domestic and import benchmarks.
• Logistics and transport benchmarks. Cost per tonne-km, port handling, regional haul rates.
• Labour cost indices. By skill grade and region.

For a contractor pricing KSA infrastructure or industrial projects, these are reference rates you would otherwise gather from disconnected supplier quotes and SAMA reports.

KAPSARC's open-data philosophy means most of these tables update monthly or quarterly, with structured CSVs and an API. That update cadence is what makes them useful as a live pricing source.

How ORKSTRA pulls KAPSARC in

ORKSTRA's pricing engine treats government data as one of four weighted sources behind every rate suggestion:

| Source | Weight | Origin | |---|---|---| | Historical | 40% | Your own past projects | | Government | 30% | KAPSARC (KSA), CAPMAS (Egypt), UAE Stats | | Crowd | 20% | Anonymised contractor pool (k-anonymity ≥ 5) | | Heuristic AI | 10% | Fallback model |

The KAPSARC integration runs on a daily 02:00 UAE-time scheduled job. It hits the open KAPSARC endpoints, respects robots.txt and rate limits, normalises the response into the internal pricing_rates table with source='govt' and tenant_id NULL, and tags each row with a freshness timestamp.

When a QS opens a BOQ line for, say, "diesel for site generators", the pricing engine queries the government source alongside the others, applies the source weighting, and surfaces a single number plus a transparency panel that shows each component.

What the QS sees

On every priced BOQ row, ORKSTRA shows:

• The aggregated suggestion (the weighted number).
• The four source values, each with its date.
• The spread (max minus min) as a confidence proxy.
• A link to drill into any single source.

This is the transparency contract. The QS is never asked to trust a black-box rate. They see what each source said, and they decide.

Where KAPSARC is strong, and where it is not

Strong

• Energy inputs. If your project burns diesel, the KAPSARC monthly average is the single most defensible reference rate.
• Bulk material indices. Cement, steel and aggregate are well covered with monthly granularity.
• Macro labour trends. Useful for index-linked contract escalation clauses.

Not so strong

• Specialty trades. Specialised joinery, custom MEP equipment, imported finishes — KAPSARC does not track these. Crowd-sourced and historical sources fill that gap.
• Sub-regional variation. KAPSARC publishes at country and broad regional level. Project-site variation against city centres is not captured.
• Real-time spikes. The monthly cadence misses week-on-week volatility for commodities. The engine flags this with a "stale if older than N days" rule and downweights stale government rates.

The drift alert

KAPSARC values drift over time. When the difference between the KAPSARC rate and the historical or crowd rate exceeds 10%, ORKSTRA raises a drift alert in the admin dashboard. That is the QS's cue to investigate before the next tender. Common causes:

• A KAPSARC dataset definitional change (rare but real).
• A genuine market move (energy spike, steel price collapse).
• A historical anomaly (one outlier project skewing the local mean).

The alert is non-blocking. It is a prompt, not a halt. The QS decides whether to recalibrate.

What happens when KAPSARC is silent

For items KAPSARC does not cover, the engine simply downweights the government source to zero and reweights the remaining three. Historical (40%) and crowd (20%) carry more weight, and the heuristic AI fallback (10%) covers the gap.

This graceful degradation matters. A QS pricing a custom marble finish should not see "no KAPSARC data — no suggestion". They should see a suggestion built from the sources that do have signal, with the missing source clearly marked.

What this is not

KAPSARC integration is not a magic compliance feature. It does not make your bid government-approved or your audit risk lower by itself. What it does is anchor your rates to a defensible external reference that you can show in a consultant meeting or a procurement audit, without arguing about whose Excel was used.

The KAPSARC rate becomes a footnote you can cite: "as of 2026-05-15, KAPSARC's published electricity unit cost for industrial use in the Eastern Province is X SAR/kWh, which the priced rate reflects."

That is the kind of small but real defensibility that compounds over 50 projects.

Where to start

If you are pricing KSA work, KAPSARC integration is one of the fastest wins available. You do not need to set anything up — it is on by default on every tenant with KSA projects flagged.

• /demo — see the KAPSARC source row light up on a KSA BOQ pricing screen.
• /premium-stack — the technical map of how government data flows into the pricing engine.

— Eng. Amr Shoieb