GNI Multi-Molecule Digital Twin

Decision-Support Use Cases

Human-led engineering scenarios. Every run is recorded in the audit log.

Reverse-compression siting
Why better decision-support is needed

Summer biomethane injection is throttled today because low-pressure distribution pockets hit MAOP with nowhere for the gas to go - curtailed producers lose revenue and Ireland misses renewable-gas targets. Siting reverse-compression by intuition or one-off desktop studies is slow and inconsistent. A repeatable, network-wide search of every constrained edge lets planners rank interventions on real relief per euro instead of anecdote.

Engineering calculations & methodology
  1. Screen every distribution edge (pipe_type = DISTRIBUTION) for stagnant or reversing flow: |v| ≤ 0.05 m/s.
  2. Retain edges whose operating pressure reaches ≥ 90% of the lower-endpoint MAOP - the hydraulic condition that forces producer curtailment.
  3. Restrict the search to the selected producer's catchment: constraints whose nearer endpoint lies within the chosen radius of the connection point (whole network when no producer is selected).
  4. Identify the low-pressure endpoint of each retained edge as the constraint location, and build the candidate set of tie-in points from all nodes carrying transmission pipe.
  5. Select the nearest transmission candidate by great-circle (haversine) distance, then compute the true build route with a Dijkstra shortest path over the network graph weighted by segment length; straight-line distance is the fallback where no path exists.
  6. Compute the producer haul length (Dijkstra from the connection point to the constraint), convert tie-in route length to annual volumetric relief, and rank candidates by total constructible length - lowest capital per GWh relieved.
constrained ⇔ |v| ≤ v_stag AND P_op ≥ 0.9 · MAOP_min(source,target)
in catchment ⇔ min(d_gc(prod, src), d_gc(prod, tgt)) ≤ r
d_gc = 2R·asin(√(sin²(Δφ/2) + cosφ₁·cosφ₂·sin²(Δλ/2)))
L_route = min Σ len(e) (Dijkstra, e ∈ path)
L_total = L_haul + L_route
Relief [GWh/yr] = (L_route / 1000) × 4.5 GWh·km⁻¹·yr⁻¹

Assumptions & limits: v_stag = 0.05 m/s · MAOP utilisation threshold 0.90 · relief intensity 4.5 GWh per km of tie-in (planning-level heuristic) · R = 6 371 km

Scans low-pressure distribution edges with stagnant/reversing flow near MAOP within the producer's catchment, then finds the shortest network route to the high-pressure transmission grid to relieve summer injection curtailment.

Highlighting a connection point marks it green on the map.

Loading map…