Chemical Tracer Flowback Data Help Understanding of Fluid Distribution

You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.

To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT

This paper presents a data set involving the pumping of multiple, unique chemical tracers into a single Wolfcamp B fracture stage. The goal of the tracer test is to improve understanding of the flowback characteristics of individually tagged fluid and sand segments by adding another layer of granularity to a typical tracer-flowback report. The added intrastage-level detail can provide insights into fracture behavior in shale-reservoir stimulation by looking at individual fluid-segment tracer recoveries.


Operators have relied upon high-­intensity completion designs that include a combination of high proppant volumes, increased perforation-cluster density, and smaller-mesh-size proppants. These designs aim to create a complex fracture network and increase the contact area with shale rock. They have helped operators achieve higher initial productivity and larger estimated ultimate recovery while simultaneously enabling the drilling of horizontal wells at tighter well spacing. The traditional, biwing fracture model seems to be scrutinized increasingly for its lack of relevance when stimulating shale reservoirs. Operators have observed greater fracture complexity when using enhanced completion designs. These designs aim to increase fracture surface area and complexity, leading to a debate regarding the merits of stimulated reservoir volume (SRV) and propped-stimulated reservoir volume, also known as effective propped volume (EPV).

SRV, estimated usually from microseismic mapping, is a rough estimate of the volume of rock that is hydraulically fractured, and is sometimes defined as the product of gross stimulated area and pay-zone thickness. EPV is a fraction of the total SRV that is supported by proppant and is capable of flowing during depletion. From a production perspective, the surface-area contact of the fractional propped SRV is more important than the gross SRV estimate.

In the past, chemical tracer data have offered stage-level insights into load recovery and hydrocarbon contribution, but the data set presented in the complete paper considers individual fluid-segment data within a single fracturing stage. A few of the questions prompting this study included:

  • Do certain fluid segments exhibit poor tracer recovery by being placed within the unpropped fraction of the SRV?
  • Does the order of injected fracturing fluid correspond with the order in which tracer is produced?
  • Can the residence-time calculation for each tracer be used to infer the degree of fracture complexity?

As operators elect to enhance fracture complexity by increasing perforation-cluster density, using lower-viscosity-fluid systems, and pumping smaller-mesh proppants, the modeling of fracture geometry has proved difficult. In addition, varying perforation-cluster efficiency and sand-duning effects can cause fluid and proppant to be distributed nonuniformly within the fracture network.

Without these effects, a piston-like displacement of the fracturing fluid is expected within a given fracture (Fig. 1). In the simplified illustration, the pad-stage fracturing fluid should flow furthest into the formation. It is also possible that a limited recovery of the pad stage is to be observed if it is placed outside of the propped fracture network. By tagging each fluid and sand segment of the fracture stage, the residence-time distribution of each tracer should yield insight into the fluid and proppant distribution of the fracture network.

Fig. 1—Piston-like displacement. Each color represents a unique tracer.

Overview of Chemical Tracers

During the past decade, chemical tracer usage for diagnosing fracture treatments in unconventional reservoirs has increased dramatically. Chemical tracers are exclusively soluble in a single target phase, which can be water, oil, or gas. They can also be manufactured in liquid or solid form. Liquid tracers are pumped with fracturing fluid at a constant concentration, which will scale proportionately with the clean fluid rate of the fracturing blender. Liquid water tracers follow the aqueous phase, whereas liquid hydrocarbon tracers dissolve into oil or gas upon contact with reservoir fluids. A new addition to tracer technology results from adsorbing chemical tracers onto a solid carrier. Solid-tracer carriers have the benefit of being placed with the proppant in the fracture network; however, the specific mass of tracer absorbed onto each carrier bead and the release rate of the tracer from the solid carrier in the reservoir is difficult to quantify. Solid tracers also degrade during the fracturing treatment when exposed to high treating pressures and do not have the crush strength of sand or manufactured proppant. Tracers can also come in polymer strip form, wherein the tracer can be attached to completion hardware such as sand-control screens.

Typical deliverables from chemical tracer tests include analysis regarding fracturing-fluid recovery, wellbore cleanup, and stage inflow contribution. Advanced research has led to use of chemical tracer flowback data to investigate fracture geometry and fracture-­network complexity. Studies using chemical tracers to evaluate fracturing-fluid distribution within a single fracture stage are limited.

Test-Site Description

This project was conducted in a horizontal well drilled and completed in the Wolfcamp B formation of the Delaware Basin. The lateral length is approximately 10,000 ft with 48 fracture stages. Each stage consisted of approximately 500,000 gal of water-based fracturing fluid and 451,000 lbm of sand. Slick­water and linear gel were used as the carrier ­fluids. 100-mesh sand was pumped from 0.2 to 1 lbm/gal, and 40/70-mesh followed from 1 to 2 lbm/gal. The entire stage treatment was divided into 15 segments on the basis of the fracturing fluid, proppant size, and concentration.

Tracer Application

Twelve unique water-phase tracers and oil-phase tracers were injected into the heel-most stage to tag individual fluid and sand segments beginning with the pad segment and ending with the 2-lbm/gal 40/70-mesh stage. Each fluid and sand segment was allocated unique tracers. The tracers were injected continuously along with the fracturing fluid at a constant concentration throughout each individual segment. The chemical tracers are conservative tracers, exclusively soluble in their carrier fluid, and will not decompose or react under subsurface conditions.

After the well was brought on line for flowback and production, water and oil samples were collected at the wellhead. One sample per day was taken for the first 2 weeks, followed by a reduction in sampling frequency. The entire sampling process lasted approximately 80 days, during which 21 water samples and 21 oil samples were collected. All samples were sent to a laboratory for analysis.

Results and Discussion

Raw Tracer Data. Tracers injected into different fluid and sand segments were commingled to some degree. Most of the tracers pumped were detected in each sample, including the earliest flowback samples, showing that some fluid pumped early in the treatment was placed near the wellbore, indicating a degree of fracture complexity. Significant connectivity between primary and secondary fractures may result in sufficient blending of fracturing-fluid segments.

In general, the tracer concentration in each fluid sample declined over time because of dilution from increasing formation-water influx. However, the tracer concentrations then increased suddenly, suggesting that the water-flow contribution from the heel stage increased relative to the previous sample date. This would normally be verified by relative changes in tracer production from other fracture stages if that data were available. The other fracture stages of this well were not traced.

Normalized Tracer Data. Normalizing raw tracer data to their injected amount accounts for the concentration variation owing solely to the different segment volumes. A segment for which 476 bbl of fracturing fluid were pumped downhole was selected as the standard, and corresponding normalization coefficients were then applied to other tracers. The earliest segments for both phases of tracer display overall increasing concentration, indicating that the earlier segments are more likely to partially flow into up-propped SRV and are at a much lower concentration relative to the other tracers during flowback and production.

Differing magnitudes of concentration imply that the tracers are contacting completely different pockets of hydrocarbon throughout the stage. This further supports the idea of increased fracture complexity with this style of completion program.

The complete paper describes the methodology of residence-time-distribution analysis.


Water-soluble and oil-soluble tracers were injected into 12 segments within a single heel stage. Flowback samples were collected and analyzed to detect tracer concentration. Application of chemical tracers is an effective, nonintrusive technique that provides invaluable information about the fracturing-fluid dynamic inside fractures. The following conclusions were reached:

  • All tracers were detected in the flowback samples, indicating that no fluid segments existed that were placed completely outside of the propped fracture network.
  • A high degree of fluid mixing appears to take place during stimulation, possibly an indication of high fracture complexity.
  • Residence-time-distribution analysis should be used to determine the transit time from the fracture to the surface for each traced segment.
  • To calculate the average residence time for each tracer, a complete tracer production history curve is needed.
  • All tracers flowed back simultaneously.
  • Tracers pumped during the linear gel fluid segments at higher sand concentration flowed back faster than the slickwater segments.
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 194362, “Understanding Fracturing-Fluid Distribution of an Individual Fracturing Stage From Chemical Tracer Flowback Data,” by Wei Tian and Alex Darnley, SPE, ResMetrics; Teddy Mohle, SPE, and Kyle Johns, Contango Oil and Gas; and Chris Dempsey, ResMetrics, prepared for the 2019 SPE Hydraulic Fracturing Technology Conference and Exhibition, The Woodlands, Texas, USA, 5–7 February. The paper has not been peer reviewed.

Chemical Tracer Flowback Data Help Understanding of Fluid Distribution

01 September 2019

Volume: 71 | Issue: 9

No editorial available



Don't miss out on the latest technology delivered to your email weekly.  Sign up for the JPT newsletter.  If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.



No editorial available