How DRA Effectiveness Modeling Helps Pipeline Operators Replace Assumptions with Data

The Challenge

Drag Reducing Agents (DRAs) are high-molecular-weight polymers used to reduce frictional pressure losses in liquid pipelines. DRA suppliers usually provide performance curves to pipeline operators, who then apply this information to their pipeline systems.

The challenge is that supplier-generated performance curves can be a great starting point, but they also might be subjective. Based on laboratory testing and, in some cases, field measurements, the performance curves may be conservative and may not fully represent the actual performance achieved under specific field conditions. 

Factors that can cause DRA effectiveness to differ from vendor expectations: 

• Differences in pipeline characteristics,
• Fluid properties,
• Pump shear, 
• Turbulence,
• Batch interfaces,
• Seasonal operating conditions. 

Considering all these factors, it is important for pipeline operators to seek an independent assessment of DRA performance. Otherwise, they pose multiple risks to a pipeline system:

• Reduced operational efficiency, 
• Excessive DRA injection and unnecessary chemical costs,
• Failure to achieve target flow rates,
• Exceeding pressure limits while attempting to meet delivery targets, 
• Repeated manual tuning of hydraulic models. 

 

The Solution

Good news is, all of these challenges could be solved if pipeline operators replace assumptions with data. Pipeline-specific DRA performance can be estimated directly from operating data, maintaining physically meaningful behavior and capturing differences between DRA products, fluids, operating conditions, and pipeline systems. 

For this purpose, Venturi designed a solution called DRA Effectiveness Modeling. It combines first-principles hydraulic calculations with machine learning and mathematical optimization. 

Here’s how it works:

• The physics-based layer uses fundamental pipeline hydraulic principals such as Reynolds number, friction factor, and static and frictional pressure losses. 
• Building on this foundation, the data-driven layer estimates field-calibrated DRA effectiveness coefficients using regression, machine learning, and optimization techniques. 

Rather than assuming a fixed mathematical relationship between DRA concentration and friction reduction, the algorithm evaluates multiple effectiveness models and automatically identifies the one that best represents the operating data.

The solution can also model DRA degradation as it travels through the pipeline. It evaluates multiple degradation models and identifies the one that best matches field measurements. This allows DRA effectiveness to be evaluated using the estimated degraded DRA concentration at each pipeline location, rather than assuming the full injected concentration remains active throughout the system.

The solution provides the following capabilities: 

• Estimates field-calibrated DRA effectiveness coefficients directly from operating data 
• Automatically identifies the most appropriate DRA effectiveness and degradation models for each pipeline 
• Generates separate DRA effectiveness coefficients by fluids, batches, seasons, and operating flow ranges 
• Allows separate calibration for low-, medium-, and high-flow operating regions 
• Supports extended effectiveness equations incorporating Reynolds number effects and additional coefficients 
• Quantifies DRA degradation due to pump shear, transport distance, and other degradation mechanisms

 

The Results

The solution has been successfully applied across multiple pipeline systems and operating environments.

For one of our clients, the model estimated field-calibrated DRA coefficients for three different products batches in less than one hour. The results closely matched manually validated values, proving that there is a potential for eliminating the extensive manual effort that was previously required on their end.

For another client, Venturi conducted the DRA degradation assessment for a long pipeline to see how the effectiveness degraded as the product traveled downstream. The pipeline included two valve stations where both pressure and temperature measurements were available. Using these measurements, the degradation model estimated the loss of active DRA effectiveness within each pipeline subsegment, allowing the client to quantify degradation rates throughout the system rather than treating the pipeline as a single uniform segment. This information provided valuable insight for optimizing injection strategies and improving hydraulic predictions over long transportation distances. 

One more interesting case involved the comparison of multiple DRA products. The operator conducted field trials and operated each product for approximately one week before sharing the data with us. After that, our algorithm independently developed field performance curves for each DRA and compared their effectiveness under identical operating conditions. The results enabled the operator to confidently select the DRA that delivered the greatest hydraulic benefit while minimizing associated chemical cost. 

For another pipeline, the solution was used to compare vendor-provided DRA performance curves with field-calibrated curves derived from operating data. Based on Figure 1, the analysis demonstrated that the vendor curves significantly underestimated the actual effectiveness of the DRA under field operating conditions, illustrating the value of calibrating DRA performance using real operating data rather than relying solely on laboratory testing.

Figure 1. Comparison of data-driven DRA effectiveness curve versus vendor-provided curve. 

In another application, the algorithm identified separate effectiveness coefficients for a DRA product designed for heavy crude but used in both heavy- and light-crude service. The calibrated models demonstrated that the DRA performed substantially better than expected in heavy crude service. More importantly, the analysis disproved the existing assumption that the DRA had negligible benefit in light crude, revealing a measurable and economically meaningful performance improvement.

 

The Bottom Line

Valuable insights extracted from operating data with the help of DRA Effectiveness Modeling enables pipeline operators to improve the way energy is moved and used. 

It replaces assumptions with validated knowledge that could immediately be applied to existing processes, leading to improved pipeline performance and reduced costs. 

 

About the Author