Optimizing Deepwater Production Systems Using Integrated Production System Modelling with CMG CoFlow

Deepwater developments involve complex interactions between reservoirs, wells, and surface facilities, where decisions made early are costly to reverse.

A study conducted with Brava Energia on the Atlanta Field in offshore Brazil applied Integrated Production System Modelling (IPSM) using CMG CoFlow, combined with uncertainty analysis and optimization.

Outcome: By modelling the full production system and optimizing key operational and design parameters, cumulative oil production increased by 21%, representing approximately 2 billion USD in additional value. 

Why Integrated Modelling Matters

Traditional workflows often:

• Model reservoir and facilities separately 
• Assume fixed bottom-hole pressures 
• Ignore system-level constraints 

This leads to:

• Misrepresentation of production potential 
• Hidden bottlenecks 
• Suboptimal operational decisions 

IPSM enables:

• Coupled simulation of Reservoirs, Wells, and Surface network 

• Dynamic interaction between subsurface and facilities 
• Realistic forecasting under operational constraints 
• Optimization of design and operational parameters 

Integrated models dynamically adjust well pressures and flow based on both reservoir behavior and facility limits, providing more accurate production forecasts. 

Operational Context: Atlanta Deepwater Asset

• Location: Santos Basin, Brazil (~1550 m water depth) 
• OOIP: ~1.8 billion barrels 
• Reservoir: High porosity (~36%) and High permeability (4-6 Darcy) 
• Fluid: Heavy oil (14° API) and High viscosity (~228 cP) 

Figure 1: Dual-reservoir system (Main and Northeast) with independent dynamics feeding into a shared production network.

The Challenge: Development Complexity

• Two independent reservoirs (Main + Northeast) 
• Shared FPSO facility and operationalizing a new FPSO
• Multiphase pumps (MPP) connecting wells in pairs 
• Strong dependence on riser-based artificial lift for viscous oil

This creates a highly coupled system in which reservoir performance and facility constraints cannot be evaluated independently.

Complex configurations, such as two wells feeding a single multiphase pump, cannot be reliably represented using conventional approaches. 

Figure 2: Integrated production system showing wells connected in pairs to multiphase pumps and a shared FPSO, requiring coupled modelling.

Solution: CMG CoFlow IPSM Workflow

CMG’s workflow combined Reservoir simulation (IMEX), Network modelling (CoFlow), and Optimization (CMOST)

1. Full System Coupling

• Reservoirs + wells + facilities solved simultaneously 
• Back-pressure and flow dynamically updated 

2. High-Fidelity Facility Modeling

• Multiphase pumps explicitly modelled 
• Flowlines, risers, and constraints fully represented 
• Switch from old FPSO to new FPSO

3. Multi-Fidelity Flexibility

• Full-physics models 
• Simplified representations where needed 

4. Integrated Optimization

• Automated scenario generation 
• Uncertainty sampling (Latin Hypercube) 
• Objective-driven optimization 

Model Overview

Reservoir Models	Surface Network & Constraints
Main Reservoir ~ 4.7M grid cells	FPSO capacity limits:  Oil: 50,000 STB/day  Liquid: 178,000 STB/day  Water: 163,000 STB/day
Northeast Reservoir ~1.3M grid cells	Pump constraints:  ΔP ≤ 180 bar  Inlet pressure ≥ 20 bar

The full network includes:

• Wells connected in pairs to multiphase pumps 
• Flowlines and risers 
• FPSO processing constraints 

Key Results
1. Bottleneck Identification (Base Case)
From simulation results:

• Early-time constraint: Oil processing capacity limits production 
• Late-time constraint: Water handling capacity becomes dominant 

Insight: As shown in Figure 3, the system transitions from being oil-rate constrained early in field life to water-handling constrained at later stages. This dynamic shift cannot be captured using static assumptions and highlights the need for integrated system modelling.

This enables operators to prioritize facility upgrades over drilling additional wells, ensuring capital is directed toward the true limiting factors in the system.

Figure 3: Time-dependent shift in system bottlenecks from oil processing to water handling capacity.

2. Well-Level Performance Diagnostics
The model identified:

• Wells shutting in due to insufficient lift 
• Specific wells requiring intervention 

Insight: Enables targeted optimization and avoids blanket field-wide changes 

Figure 4: Example of well shut-in due to insufficient lift, highlighting the need for system-level optimization.

3. High-Fidelity Modelling vs Simplified Approaches
From the CoFlow comparison:

• Traditional VFP-based approaches cannot capture pump-well interaction 
• CoFlow MPP approach fully represents facility design and eliminates the need for low-fidelity approximations 

Key Insight

Accurate system representation requires modelling actual equipment behavior, not simplified performance curves.

Figure 5: High-fidelity representation of multiphase pump systems using CoFlow, enabling accurate modelling of well-facility interaction

4. Optimization Results
Using integrated optimization:

• ~72 full IPSM simulations with 2 reservoirs and shared facilities
• Only ~12% increase in simulation time vs standalone

Figure 6: Optimized strategy delivers a significant increase in cumulative oil production compared to base case.

Key Takeaways

• 21% increase in cumulative oil production equivalent to ~28 million barrels
• ~2 billion USD value 
• Key Optimization Drivers

• Riser diameter: design decisions can have a larger impact than well count, emphasizing the importance of system-level optimization over traditional development approaches focused solely on subsurface expansion.
• FPSO water capacity: increasing water handling capacity may unlock additional production, particularly during later stages of field life when water rates dominate system constraints.

Figure 7: Sensitivity analysis identifying riser diameter and FPSO water capacity as key drivers of production performance.

Best Practices

• Use integrated models to capture reservoir–facility interactions 
• Avoid relying solely on simplified well performance curves 
• Incorporate facility constraints early in field development 
• Apply optimization workflows to: 
  • Identify key drivers 
  • Prioritize interventions 

Conclusion

This study demonstrates that:

Production optimization in complex offshore fields requires a fully integrated system perspective.

CMG CoFlow enables engineers to move from isolated component analysis to system-level decision optimization.