midstream

Skill: Midstream Pipeline

Guide AI agents through pipeline transportation and storage operations.

Purpose

Support pipeline engineers in ensuring safe and efficient hydrocarbon transportation and storage.

Roles

• Pipeline Engineer - Pipeline design, hydraulics, materials
• Operations Manager - Daily operations, scheduling, nominations
• Integrity Engineer - Inspection, corrosion management, repairs

Data Types

Data	Source	Frequency
Flow rates	SCADA	Real-time (1-min)
Pressure	SCADA	Real-time (1-min)
Temperature	SCADA	Real-time
ILI (Inline Inspection)	Smart pig	5-year cycle
Leak detection	CPM/DDS	Real-time (continuous)
Product quality	Lab	Batch/sample

Workflow

Phase 1: Transport Scheduling

1. Receive nominations (shipper requests)
2. Check pipeline capacity
3. Schedule batches/products
4. Optimize pump/compressor runs
5. Confirm deliveries

Phase 2: Real-Time Monitoring

1. Monitor pressure/flow at key points
2. Track batch locations
3. Detect anomalies (leaks, theft)
4. Adjust operations as needed
5. Communicate with control center

Phase 3: Integrity Management

1. Plan ILI (smart pigging)
2. Analyze inspection data
3. Identify defects (corrosion, cracks, dents)
4. Prioritize repairs
5. Verify repairs

Phase 4: Leak Detection

1. Monitor CPM (Computational Pipeline Monitoring)
2. Analyze mass balance
3. Check pressure/flow deviations
4. Investigate anomalies
5. Emergency response if confirmed

Domain Tasks

These tasks are handled by this skill:

• Pipeline hydraulics calculation
• Leak detection analysis
• ILI data interpretation
• Integrity assessment

Software Tasks

These tasks invoke petropowers:oil-gas-delegation:

• SCADA monitoring dashboard
• Pipeline scheduling system
• ILI data management platform
• Leak detection alerting system

Example Workflows

Pipeline Hydraulics

import math

def pressure_drop_liquid(flow_bpd, diameter_in, length_ft, viscosity_cp, density_api):
    """Calculate pressure drop for liquid pipeline (simplified)"""
    
    # Convert units
    q_bpm = flow_bpd / 1440  # bpd to bpm
    d_ft = diameter_in / 12
    
    # Velocity (ft/s)
    area_sqft = math.pi * d_ft**2 / 4
    velocity = q_bpm / 7.48 / area_sqft  # ft/s
    
    # Reynolds number
    rho = 62.4 * (141.5 / (density_api + 131.5))  # lb/ft³
    re = rho * velocity * d_ft / (viscosity_cp * 0.000672)
    
    # Friction factor (Blasius for turbulent)
    f = 0.079 / re**0.25 if re > 4000 else 64 / re
    
    # Pressure drop (psi)
    delta_p = 2 * f * (length_ft / d_ft) * rho * velocity**2 / (32.2 * 144)
    
    return delta_p

# Example: 12" pipeline, 50,000 bpd, 100 miles
d_psi = pressure_drop_liquid(
    flow_bpd=50000,
    diameter_in=12,
    length_ft=100 * 5280,
    viscosity_cp=5,
    density_api=35
)

print(f"Pressure drop: {d_psi:.0f} psi")