Decision Brief
Solve the base case to begin
READY
Active Project
100-mile Transmission Demo
Operations Overview
What can the system do now?
See current hydraulic capability, operating margin, system health, and the latest scenario without opening a detailed study.
Capacity Lens
Fast questions
GHDE Decision Architecture
Observe → Understand → Improve
PRODUCT DIRECTION
Observed operating reality and creative design/optimization are related but distinct studies. Operating practices can inform decisions without becoming unnecessary solver rules.
Network
Pressure profile
Operations Intelligence
Health monitor
Meeting Mode
Quick Decision
Change one operational condition, compare against the base case, and get an immediate decision-oriented summary.
Scenario Builder
What changed?
Plant trip is evaluated as a new steady-state endpoint in v0.2. Transient propagation and linepack timing are future physics modules.
Answer
No scenario evaluated
READY
Hydraulic margin
Energy / cost
Confidence / refinement
What-if Workspace
Scenario Lab
Build, save, rerun, and compare operating cases without changing the base model.
Saved scenarios
Scenario editor
Base vs Scenario
Comparison
v0.4.4 Decision Intelligence
Capacity & Constraint Intelligence
Quantify deliverability, identify the first hydraulic constraint, and screen which operating or network changes have the most hydraulic leverage before detailed economics or project engineering.
Question 1
How much more can we deliver?
v0.4.4 stops capacity at the first configured operating/model boundary: minimum pressure, maximum pressure where configured, operating velocity, Mach/model review, or solver convergence. The velocity advisory is a warning only. This remains steady-state screening, not a transient or contractual capacity guarantee.
Capacity Answer
Not evaluated
READY
Question 2
What pressure support is required?
Pressure Strategy
Not evaluated
READY
Constraint Ranking
What limits the system first?
v0.4.4 Capacity Profile Reliability
How does usable operating margin approach the boundary?
Hover a scenario point for pressure, velocity, Mach estimate, limiting asset, and solve status.
The upper profile shows pressure margin; the lower profile shows velocity utilization. The numerical table below is the auditable source for every plotted scenario point.
Capacity Boundary Framework
What actually stops the case?
Run a capacity search to identify the limiting condition.
v0.4.7 Constraint Foundation
Hard constraints vs advisory preferences
READY
Hard constraints reject an unacceptable candidate state. Advisory constraints flag or penalize a state without pretending the real pipeline has stopped functioning. Future optimization will use this distinction for equipment envelopes, commercial obligations and operating preferences.
Auditable Profile Data
Capacity scenario points
Build the capacity profile to review the numerical scenario points.
Low-Hanging-Fruit Screen
Which changes have the most hydraulic leverage?
The v0.4 screen perturbs one variable at a time and re-solves the network. It ranks hydraulic leverage only; CAPEX, compressor maps, tariffs, reliability and full economics remain future inputs.
v0.5 Economic Lens
What is the hydraulic opportunity worth?
v0.5 connects solved supply and delivery flows to screening supply cost, customer value, compressor-station energy and variable operating cost. Capital, tax, tariffs, reliability and detailed compressor maps remain later refinements.
v0.5 Economic Decision Baseline
Economics
Connect hydraulic feasibility to screening economics without turning GHDE into an accounting or detailed compressor-design package.
Base Case
Daily economic picture
READY
Latest Scenario
Incremental economics
NO SCENARIO
Evaluate a hydraulic scenario to compare economic consequences.
Economic Bridge
Where does daily contribution come from?
Screening economics only. Revenue/value and cost assumptions are illustrative demo inputs until replaced with project-specific commercial data.
Global Assumptions
Economic basis
ENGINEER / ANALYST INPUT
Supply Sources
Availability & tiered source economics
Whole-number flow limits are intentional. Marginal commodity costs retain decimal precision. Tiered cost curves are active in screening economics and prepared for optimization dispatch.
Customers / Deliveries
Firm, flexible, interruptible & optional service
Customers may have minimum, firm, target and maximum quantities plus competitor availability. Tiered marginal value is a decision input, not a custody-transfer billing calculation.
Compression & Energy
Electric, fuel-gas or mixed station energy
SIMPLIFIED
Station power remains piecewise-linear. Energy may be electric, fuel gas, or mixed. Fuel-gas rate uses a screening PWL curve behind the model; detailed compressor thermodynamics and unit maps remain later fidelity.
Facility / Station Foundation
Supply-location fidelity without plant simulation
FOUNDATION
Facilities are containers for supply availability, compression, energy, metering and future operating states at a network node. Detailed production-plant simulation remains intentionally outside the hydraulic network core.
v0.6 Optimization Foundation
Search feasible operating states
v0.6 can search fixed-pressure supply setpoints, reject candidates that violate hard constraints, penalize soft operating preferences, and rank feasible states by screening contribution. Detailed compressor maps and continuous nonlinear optimization remain later refinements.
v0.6.3.6 Business-Mode Optimization & Transport Corrections
Optimization
Define the goal, decision variables, hard limits, soft preferences and piecewise-linear relationships before choosing how to solve the problem. GHDE keeps the structured model visible and auditable.
Guided hybrid optimization: GHDE now classifies the configured problem and automatically orchestrates structured screening, discrete-state enumeration, deterministic global sampling, hydraulic feasibility checks, local refinement and fallback search. The deterministic hydraulic solver remains the feasibility authority.
Guided Setup
What are you trying to accomplish?
NOT CHECKED
Setup Health: checks effective executable search freedom after business-mode and operating-policy restrictions; it does not silently change the optimization problem.
Setup Health
Problem formulation & searchability diagnostics
Run Setup Health before solving. GHDE now checks whether the optimizer actually has useful executable degrees of freedom.
AUTO Solver Planner
How GHDE plans to solve it
Pre-Optimization Intelligence
Opportunity Screening & Lever Intelligence
NOT SCREENED
Before expanding the search, GHDE looks for obvious economic/hydraulic levers and for setup choices that can consume the search budget before a feasible starting point is tested.
Run the base hydraulic case, then screen opportunities.
Meaning: Green = strong current-operation lever; Yellow = plausible lever with interactions; Blue = context/scenario/design opportunity; Gray = unavailable/maintenance. The score is a heuristic prioritization, not an optimized result.
Business / Operating Model
Choose what the optimizer is actually allowed to optimize
MODE
Important: Interstate / Tariff Transport is a modeling mode, not a FERC compliance determination. Actual tariff, nomination, certificate, contract, priority, fuel-retention and operating rules must be represented for the specific pipeline.
Operating Policy
Operational constraints vs reference freedom
AS OPERATED
Reference freedom keeps physical limits and maintenance/unavailable equipment, but relaxes policies such as must-run and must-take. Full availability is a what-if benchmark, not an achievable operating recommendation.
Run more than one regime to compare the economic burden of operating restrictions.
Decision Variables
Optimization Model Builder
Supply pressure and customer-delivery variables can participate directly. Compressor integer/binary states and supported parallel-pipe alternatives participate in the hybrid search. Supply-source flow remains a dependent hydraulic outcome checked against availability and capacity.
Constraints
Hard boundaries and soft preferences
Hard constraints reject an unacceptable candidate. Soft constraints create a penalty or ranking preference. Operating practice should not become a hard rule unless there is a physical, contractual, regulatory or explicit business reason.
Piecewise-Linear Relationships
Equipment, supply and economic curves
Enter PWL points as
x:y. Flow breakpoints display as whole numbers; unit costs/values and power retain useful decimals. Supply cost, customer value and compressor-energy curves are exposed for auditable optimization setup.Current Executable Search
Supply pressure search
ACTIVE v0.6.3 HYBRID
This compatibility table controls supply-pressure bounds. The Model Builder above controls the broader hybrid decision space.
Objective & Preferences
Economic objective with soft penalties
Executable objective = daily economic contribution − soft-constraint penalties − configured discrete/design daily screening costs. Hard hydraulic, source-capacity, customer-service and compressor-mode limits reject candidates.
Calculation Runtime
Optimization progress & search confidence
IDLE
Ready. Long optimization runs execute away from the browser UI thread so the page can remain responsive.
AUTO Pre-flight
Local-minimum risk & recommended search effort
NOT ASSESSED
Advanced Configuration
Guardrails that can be unlocked
RECOMMENDED AUTO
Recommended AUTO: GHDE selects broad screening and fallback behavior from the problem structure. Switch to Advanced or Expert only when you have a reason to override the planner.
Advanced and Expert modes do not disable hydraulic hard constraints. They expose search-effort and stopping controls. Restore controls provide a quick way back if an experimental setup behaves poorly.
Recommended Candidate
Best feasible state found
NOT RUN
Run the hybrid optimizer to rank feasible operating and selected design states.
Candidate Ranking
Top feasible operating states
Rejected Candidates
Why candidates were rejected
No optimization run yet.
The explanation layer is deliberate: rejected states should identify the violated hard constraint rather than disappear behind procedural rules.
Explainable Solver
Hybrid orchestration trace
ACTIVE
Run optimization to view the solver trace.
Smart Operations
Operations Advisor
Synthesizes the hydraulic result, latest scenario, model confidence, and Pipeline Health Monitor evidence into a short list of items worth attention. It explains rather than replaces engineering judgment.
Priority actions
Refinement triggers
Hydraulic Lens
Constraint ranking
v0.6.2.1 GIS Reliability
Network Map
Geographic hydraulic decision view for screening and scenarios.
GIS FOUNDATION
GIS-aware, not enterprise GIS. Geographic route vertices stay separate from hydraulic topology.
Map controls
Hydraulic network geography
READY
Use
assetId, or a feature name matching a GHDE node/pipe ID. KMZ and Shapefile are deferred.Development basemap uses MapLibre demo tiles. Select an approved provider for production customer deployments.
Selected asset
Engineering readout
Click a node or pipeline.
Geometry model
Map vs hydraulics
Map geometry does not change hydraulic length, diameter, roughness or topology.
Base Case
Network Model
Edit nodes and pipes used by the steady-state solver. The base case is preserved when scenarios are evaluated.
System View
Hydraulic network schematic
Schematic layers
Arrows follow solved flow sign, not pipe From → To definition. Near-zero flow has no arrow.
Solve the base case to populate hydraulic results.
fixed-pressure supply
demand / junction
enabled pipe
unavailable pipe
pressure margin warning
Schematic only — not geographic/GIS scale. Mouse wheel zooms around the pointer; drag blank schematic space to pan. Text Tilt rotates node names/IDs (pressures stay horizontal). Auto label density hides secondary labels at Fit View and reveals more as you zoom in. Fit View resets zoom/pan.
Nodes
X/Y are relative schematic coordinates, not GIS coordinates. Any finite values including negatives are valid; only relative spacing matters and Fit View scales the complete network. Leave blank to use automatic placement.
| ID | Name | Fixed P | P kPa abs | Demand Sm³/h | Min P | Max P | Elevation m | Station/Mile | X | Y |
|---|
Pipes
Parallel pipes are offset in the schematic and remain individually selectable.
| ID | From | To | Length km | ID mm | Rough mm | Loss × | Enabled |
|---|
Fast Automatic Review
Network feasibility & red flags
NOT RUN
First-pass screening before optimization: convergence, pressure, velocity, supply/interconnect capacity and topology. It highlights where a human or later AI/optimization layer should investigate.
Facility Equipment
Compressor station configuration catalog
SCREENING MODEL
Load the large benchmark to review station combinations.
Unit combinations and piecewise energy/capacity curves support economics and hybrid optimization. Detailed inline suction-to-discharge compressor hydraulics are a future solver element.
Decision Workflow
Feasibility → scenarios → optimization → AI oversight
FOUNDATION
Solver Model
Boundary conditions, constraints & pipe controls
Fixed P = Yes
Known pressure boundary
The node pressure is prescribed. The network determines the resulting supply/withdrawal flow needed to satisfy connected demands.
Fixed P = No
Pressure is solved
Demand is prescribed and Newton/network continuity solves the node pressure. The entered P is an initial guess, not a control rule.
Min / Max P
Operating constraints
These screen the solved condition and capacity scenarios. They do not add procedural rules to the hydraulic equations.
Loss × / Enabled
Resistance / topology
Loss × multiplies Darcy friction loss (1.0 nominal). Enabled = No removes the pipe from the active hydraulic network for outage/routing studies.
GHDE intentionally favors boundary conditions + physical/operating constraints over a large procedural rule engine. Future compressors, regulators and optimization layers should add equipment envelopes and objective/constraint logic rather than force unnecessary textbook operating rules into the network solve.
Selected Connection
Parallel pipe comparison
Select a pipe in the schematic to compare all pipes connecting the same two nodes.
Historian Screening
SCADA Data Conditioning
Review historian data before using it to judge the pipeline. Validate timestamps, units, quality flags, known maintenance/calibration periods, and steady-state suitability before model reconciliation.
1 · Import
Historian CSV
Preferred timestamp:
YYYY-MM-DDTHH:MM:SSZ (UTC/GMT). The importer accepts the new long historian format and the simpler wide sample format. Imported data are reviewed first; they are not automatically used to calibrate the hydraulic model.2 · Quality Gate
Import checks
Checks include invalid timestamps, nonnumeric values, unsupported units, duplicate tag/timestamp records, SCADA quality flags, and explicit source exclusions.
3 · Filter
Choose data worth analyzing
Known Context
Exclude known bad/maintenance periods
This is intentionally simple. Later versions can import maintenance logs, calibration records, alarms, and historian annotations automatically.
Tag Catalog
What did we import?
4 · Steady-State Screen
Find candidate stable periods
Screening logic: a window passes when the included pressure, temperature, and flow tags remain within the engineer-set ranges. This identifies candidate steady periods; it does not prove thermodynamic or operational steady state.
5 · Visual Review
Data Quality & Steady-State Timeline
Import data and choose a tag to inspect the timeline.
Included / good
BAD quality
Candidate steady-state window
Manual/source exclusion
The timeline is a visual screening aid. Candidate windows are generated from the engineer-set stability limits above. Review operating context before accepting a window for hydraulic reconciliation.
6 · Candidate Windows
Select steady periods for later model evaluation
Selection is advisory in v0.4.2: accepted windows are saved with the local project, but automatic tag-to-asset mapping and hydraulic snapshot creation remain planned capabilities.
7 · Review Before Model Use
Filtered data preview
Future integration placeholder: tag-to-node/pipe mapping and automatic creation of hydraulic trust snapshots will be added after the data-conditioning workflow is proven. v0.3D deliberately keeps raw historian screening separate from model calibration.
Diagnose
Pipeline Health Monitor
This is a separate but complementary capability: the hydraulic solver predicts expected behavior; the health monitor asks why measured behavior differs from expected.
Health test scenario
Recommended next checks
Issue family ranking
Segment evidence
Why the engine said this
Model Reconciliation
Hydraulic Trust Baseline
Separate physical calibration from data-quality problems. Compare steady-state SCADA observations with the hydraulic model, identify suspicious instruments or asset data, and apply controlled pipe-loss multipliers only when the evidence supports it.
Historical Snapshot
Measured vs expected
Calibration fits section-level Darcy loss multipliers against measured pressures. A factor far from 1.0 is evidence to investigate—not permission to hide bad pipe data, a pinched valve, or an instrument problem.
Controlled Calibration
Pipe hydraulic factors
No calibration has been run.
Pressure residuals
Data-quality & physical hypotheses
Measurement Confidence
Uncertainty analysis
Suggested values are screening assumptions only—not a calibration certificate, custody-transfer compliance statement, or NIST-traceable uncertainty budget. Engineers can replace them with site standards or certificate values.
Decision Significance
Residual vs expected uncertainty
This answers a practical question: is measured/model disagreement small relative to expected measurement uncertainty, or large enough to justify investigation?
Historian Foundation
SCADA history & residual persistence
v0.3D: raw historian review now belongs in SCADA Data Prep. The legacy residual-series import remains here for backward compatibility after measured/model values already exist. Tag-to-asset mapping, automated model solves, drift/change-point detection, and Azure historian connectors remain future expansions.
Across Operating Conditions
Historical consistency matrix
A correction that appears in only one operating snapshot may be an event or instrument issue. A repeated hydraulic signature across stable operating conditions is stronger evidence for a persistent model/asset correction.
Asset Data
Model plausibility checks
Regression Guard
Solver self-tests
Historical Repetition
Calibration consistency across snapshots
A correction factor that repeats across independent stable operating periods is more credible than one inferred from a single snapshot. Mixed fault scenarios should not be expected to agree.
Interpretation
Trust rules in v0.2
Instrument first when isolated.A single pressure or flow residual with otherwise coherent hydraulics should trigger tag/scaling/calibration checks before hydraulic tuning.
Physical restriction when local loss persists.A repeated need for higher loss across one section can represent a pinched valve, fouling, smaller ID, roughness, fittings, or equivalent-length error.
Calibration is bounded and visible.Section multipliers remain explicit model parameters with audit meaning; they are not silently embedded in the equation.
Historical repetition matters.A factor inferred from one snapshot is preliminary. Confidence increases when the same correction explains multiple stable operating periods.
Engineering Detail
Hydraulic Solver Results
Inspect convergence, pressure, flow, velocity, Reynolds number, friction factor, density, and Z used by the current calculation.
Node results
Pipe results
Solver iteration log
v0.3 Engineering Authority
Trusted Hydraulic Core
Run permanent core checks for convergence, elevation/static head, gas-property sanity, and legacy-vs-v0.3 comparison. These are regression checks, not a substitute for independent field validation.
Core regression checks
Trusted-core scope
Hydraulic formulationIsothermal Darcy-Weisbach, integrated P²
ElevationActive at node level
Natural-gas viscosityLee-Gonzalez-Eakin
Network solutionFinite-difference Newton + damping
Property reference pathREFPROP / NIST AGA8 validation workflow
Important: AGA8 is not silently approximated in the browser. v0.3 establishes the validation interface and references the official NIST AGA8 code; an authoritative runtime AGA8 implementation belongs in the future backend/property service.
Reference-validation status
The NIST AGA8 repository is the designated AGA8 reference implementation for future comparison cases. Your licensed REFPROP installation remains the preferred independent property reference for the local validation matrix.
v0.3 deliberately separates implemented physics from reference methods so the application never reports an AGA8 result unless an AGA8 calculation was actually performed.
v0.3B.1 Engineering Validation
REFPROP Property Validation
Import the combined REFPROP 9.1 CSV generated in Windows/Parallels and compare the same pressure, temperature, and composition points against GHDE runtime property models. Reference data never changes the hydraulic solution.
Reference dataset
REFERENCE ONLY
No REFPROP CSV loaded.
Design rule: imported REFPROP values are an independent validation authority. They are not substituted into the GHDE network solver and do not tune model coefficients automatically.
Model applicability
Load a reference dataset to see applicability.
Accuracy matrix
Import a REFPROP CSV to calculate model errors.
Z vs pressure
Hover over the chart for nearest-point values.
Z error vs pressure
Hover over the chart for nearest-point values. Error = (GHDE − REFPROP) / REFPROP × 100.
Density error vs pressure
Hover over the chart for nearest-point values.
Phase / reference coverage
Single-phase boundary: REFPROP may return valid two-phase or liquid-side states that are outside GHDE's current gas-hydraulic solver. Those points are retained as reference data but excluded from the single-phase error metrics.
Largest Z discrepancies
No comparison loaded.
Thermodynamics
Gas Properties & Validation
Keep the gas-property layer visible and testable while the network solver stays model-agnostic.
Composition · mole %
Property diagnostics
v0.2 confidence: DAK/CNGA are SG-based correlations. PR is pure-component only. AGA8/GERG and mixture PR are planned validation upgrades.
Z vs pressure
Z-factor heat map
Traceability
Run History
Every scenario evaluation is stored locally in v0.2 with the model version, gas model, result, and timestamp.
Recent runs
Display Layer
Engineering Units & Standard Conditions
The solver never changes unit systems. Its canonical calculations remain kPa absolute, kelvin, metres, and standard m³/h at the project core base conditions. These settings convert only user-facing values.
Customer Demonstrations
Load a coherent example system
DEMO DATA
International / SI demo
160-km fictional transmission example using bar(g), km, mm, °C and standard m³/h.
160-km fictional transmission example using bar(g), km, mm, °C and standard m³/h.
U.S. pipeline demo
100-mile fictional transmission example using psig, miles, inches, °F and MMSCFD.
100-mile fictional transmission example using psig, miles, inches, °F and MMSCFD.
The large benchmark is a fictional ~600-mile multi-supply transmission system for outage, debottleneck, new-customer, feasibility and hybrid-optimization studies. Demonstration data only.
Unit profile
Reset restores the known-good SI Engineering display profile. It does not change the solver core or project data.
Core Invariant
Calculation units never change
LOCKED SI CORE
Design rule: unit selections are converted at the UI boundary. No alternate unit system enters the Newton solver or gas-property equations. This sharply reduces unit-conversion defects.
Conversion preview
Standard-volume conversion preserves molar flow between the project core base state and the selected reporting base state. The current gas-property model is used for the small Z correction at both base states when available.
Assumptions
Settings
These controls define base conditions and preliminary decision thresholds. They should later become project-level controlled inputs.
Gas / standard conditions
Operational thresholds
Hydraulic core
Use the v0.2 formulation only for regression comparison. New projects should use the v0.3 core.
Preliminary compressor-cost estimator
The energy estimator is intentionally separate from the hydraulic solver and is labeled preliminary until compressor maps and station equipment are modeled.