See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/341941170

Soil Mechanics and Calculating Hydraulic Fracture Risk

Conference Paper · June 2020

CITATIONS READS
0 201

2 authors:

Mark A Miller Jonathan L. Robison

GeoEngineers, Inc. GeoEngineers, Inc.
3 PUBLICATIONS   3 CITATIONS    17 PUBLICATIONS   12 CITATIONS   

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Jonathan L. Robison on 05 June 2020.

The user has requested enhancement of the downloaded file.

 North American Society for Trenchless Technology (NASTT)
NASTT’s 2018 No-Dig Show

Chicago, Illinois
March 18-21, 2019

TA-T2-03

Soil Mechanics and Calculating Hydraulic Fracture Risk

Mark A. Miller, PE, GeoEngineers, Inc., Springfield, Missouri
Jonathan L. Robison, PE, GeoEngineers, Inc., Springfield, Missouri

1. ABSTRACT

Geotechnical and trenchless engineers are commonly asked to assess the risk of hydraulic fracture and inadvertent
drilling fluid surface returns (IR) for horizontal directional drilling (HDD) and Direct Pipe® (DP) projects. One of the
most commonly used quantitative methods uses cavity expansion theory and the Delft equation to produce a Formation
Limit Pressure (FLP). By comparing the FLP to the estimated drilling fluid pressure at any given point along an HDD
profile, a factor of safety (FOS) against hydraulic fracture can be estimated. Over roughly the last decade, these
assessments have become widely used in the industry and increasingly reliable as a predictor of hydraulic fracture and
subsequent IR risk in soil, if:

• Good geotechnical data is available and properly interpreted;

• The engineer’s assumptions of tooling and other inputs are reasonable and/or calibrated to the contractor’s
actual construction plan; and
• A rigorous monitoring program is implemented during construction.

As a practical consideration, as with any calculation, the results of the cavity expansion equations are only as good as
the input parameters used. Recent, high-profile failures to use appropriate soil properties during the design process
have resulted in permit violations and very large construction claims. This paper will provide a review of soil
mechanics principles, including: drained and undrained shear strength; the results of various field and laboratory
strength tests; total and effective stress; the effect of groundwater table assumptions; and other important geotechnical
considerations for proper implementation of the Delft equation.

2. INTRODUCTION

During HDD pilot hole operations, drilling fluid (consisting typically of a mixture of bentonite and water), is
introduced at the leading end of the HDD drill stem at pressures sufficient to maintain flow and carry cuttings back to
the entry point through the annulus of the hole. As the pilot hole progresses, the annular pressures must increase to
maintain circulation over increasing distances as a result of frictional losses. In addition to the pressure and flow
volumes required to maintain circulation, hydrostatic pressure is a significant contributor to the total pressure
downhole, particularly when the HDD profile is relatively deep.

The DP construction method incorporates two, independent fluid systems: the drilling fluid and cuttings return system
that balances earth and groundwater pressures at the face of the machine and carries cuttings back to the surface; and
the lubrication fluid system that fills the annular space created by the cutting wheel overcut with bentonite-based
lubrication fluid to reduce the friction installation forces and stabilize the excavation walls. The DP system typically
operates at much lower pressures than HDD, but not necessarily insignificant pressures relative to the soil FLP
(Robison and Sparks, 2015).

Paper TA-T2-03 - 1
In soil formations, if the fluid pressure exceeds the FLP, then hydraulic fracture of the soil is likely to occur, and a
drilling fluid surface release may result. This paper will focus on the assessment of the FLP. Methods to estimate the
downhole pressure for HDD and DP during design and to monitor it during construction are found in many other
resources, such as Bourgoyne, et al (1991), and the previously noted reference, and will not be covered here.

The purpose of this paper is to provide a basis of understanding of soil mechanics principles as they relate to one of
the most commonly used assessments of hydraulic fracture risk in soil, the Delft method as described in Appendix B
of the United States Army Corps of Engineers Report CPAR-GL-98 (Staheli, et al., 1998), and elsewhere. The
common industry approach to assessing the risk of hydraulic fracture for trenchless pipeline installations is to divide
the estimated FLP of the geologic formation the installation is passing through by the anticipated downhole annular
pressures generated during construction with the result being considered the “factor of safety” against hydraulic
fracture of the geologic formation at depth.

The Delft equation assesses the ability of soil to resist hydraulic fracture and is fundamentally a geotechnical
engineering calculation for which an understanding of soil mechanics is necessary. In the author’s experience, the
input parameters of the Delft equation are too often inadequately understood and incorrectly applied, resulting in the
mischaracterization of risk and, sometimes significant and unanticipated hydraulic fracture and inadvertent drilling
fluid releases.

The Delft equation for the maximum formation limit pressure is shown in Equation 1.

−𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠
2 1+𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠
𝑅𝑅𝑜𝑜 𝜎𝜎𝑜𝑜′ ∘ 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 + 𝑐𝑐 ∘ 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
𝑃𝑃𝑚𝑚𝑚𝑚𝑚𝑚 = 𝑢𝑢 + (𝜎𝜎𝑜𝑜′ ∘ (1 + 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠) + 𝑐𝑐 ∘ 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 + 𝑐𝑐 ∘ 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐) ∘ �� � + � �� − 𝑐𝑐 ∘ 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 [1]
𝑅𝑅𝑝𝑝 𝐺𝐺

Where:
Pmax Formation Limit Pressure – Theoretical maximum pressure or total stress the soil can withstand before
hydraulic fracture occurs
φ Angle of internal friction, degrees
c Cohesion, psf
σ’0 Effective vertical stress, psf
G Shear Modulus, psf
R0 Radius of drilled hole, ft
Rp Maximum radius of plastic deformation of drilled hole, ft
u initial pore water pressure (psf)

Below we will provide a geotechnical engineering basis for understanding the input parameters and applying the Delft
equation. We do not have the space to treat each subject rigorously, only summaries can be provided, and the reader
should allow for exceptions and the use of engineering judgment, as with any geotechnical engineering practice.

3. SOIL MECHANICAL PROPERTIES

3.1 Soil Characterization and Composition

All soils are composed of soil particles and void spaces which (typically) may be air or water filled, or a combination
of air and water depending upon the degree of soil saturation. Coarse-grained soils are composed of particles with
more than 50 percent material retained on the number 200 sieve (200 openings per inch); these include sands, gravels
and coarser materials. Fine-grained soils are composed of particles with more than 50 percent passing the number
200 sieve, such as silts and clays. Considering the individual particle size and porosity of the soil types, the finer-
grained the soil and the lower its hydraulic conductivity (or permeability), the less able saturated soils are to rapidly

Paper TA-T2-03 - 2
release pore water upon loading. In nature, uniformly graded soils are rare; we usually encounter soils with multiple
constituents of various shapes and sizes and the geotechnical engineer must judge what component of the soil matrix
dominates or controls its engineering behavior. In the interest of brevity, for the purposes of this paper we will consider
soils to be saturated and either fine- or coarse-grained. Example grain size distributions for four different sands are
shown below in Figure 1.

Figure 1. Example grain size distribution (Sand).

3.2 Effective and Total Stress in Soil

Total stress below the ground surface is calculated by taking the average weight of the soil and water above the point
of interest, multiplied by the height of the soil and water column. Effective stress is defined as total stress minus the
hydrostatic groundwater pressure. It can be useful to compare the total stress at the HDD profile depth to the results
of the anticipated downhole pressures during construction and the formational limit pressure calculations; this
comparison is a good indicator of the level of conservatism in the overall hydraulic fracture risk analysis.

3.3 Soil Strength

In simplified terms, the shear strength of a soil is the limiting load or pressure it can tolerate before excessive
deformation occurs. In civil works, two types of soil loadings are commonly considered: rapid and long-term. Rapid
loads occur relatively quickly (loading develops instantaneously or within minutes or hours) while long-term loads
may develop over the course of days, weeks, months, or longer. For the purposes of hydraulic fracture assessment, the

Paper TA-T2-03 - 3
loading condition should always be assumed to develop rapidly as the soil has relatively little time to adjust to the
stresses imposed on it during the drilling process.

Under rapid loading conditions, soil shear strength is commonly considered in two cases: drained and undrained
conditions, depending upon soil type. Drained condition can be assumed for coarse grained soils with low to moderate
fines contents that allow relatively rapid flow of pore water out of them upon loading, due to their higher porosities.
Undrained condition should be assumed for rapid loading of fine-grained soils. Soil response to loading behavior is
described in general terms by the Mohr’s circle illustrations below in Figure 2. A coarse-grained (cohesionless) soil
failure envelope assuming drained conditions is represented by the angled line with a y-intercept of 0 and slope angle
φ. A fine-grained soil failure envelope assuming undrained conditions is represented by a y-intercept of c and slope
angle φ of 0.

Figure 2. Illustration of shear strength parameters using Mohr’s Circle- Terzhagi and Peck (1967).

This concept may be further described more simply as, coarse-grained soils can increase in strength with loading as
water is forced out of the void spaces and the particles reorient themselves into denser and stronger configurations;
whereas fine-grained soils do not have the ability to rapidly drain nor gain significant strength with rapid loading and
begin to shear (fail) when the limiting condition is reached.

Relating shear strength back to hydraulic fracture risk assessment: if drained conditions are (improperly) assumed for
cohesive soils, the Delft Equation will overestimate the formational limit pressure, sometimes very significantly,
indicating potentially low risk of hydraulic fracture when higher risk is in fact present. This, unfortunately, occurs in
practice too often and has led to some high-profile construction permit violations.

3.3.1 Soil Strength Application to Hydraulic Fracture Calculation

Applying the soil shear strength concepts to the Delft equation, consider the following:

1. Coarse-grained soils will have a φ angle that can be estimated from field and laboratory testing data. Coarse-
grained soils should be considered to have a c of 0. So, for a coarse-grained soil, the Delft equation for total
allowable stress reduces to (equation 2):
−𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠
2 1+𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠
𝑅𝑅𝑜𝑜 𝜎𝜎𝑜𝑜′ ∘𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠
𝑃𝑃𝑚𝑚𝑚𝑚𝑚𝑚 = 𝑢𝑢 + (𝜎𝜎𝑜𝑜′ ∘ (1 + 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠)) ∘ �� � + � �� [2]
𝑅𝑅𝑝𝑝 𝐺𝐺

Paper TA-T2-03 - 4
 2. Fine-grained soils will have a c estimated from field and laboratory testing data. There should be no
φ assigned to fine-grained soils for the rapid loading hydraulic fracture scenario. So, for a fine-grained soil,
the Delft equation for total allowable stress reduces to (equation 3):

𝑃𝑃𝑚𝑚𝑚𝑚𝑚𝑚 = 𝑢𝑢 + 𝜎𝜎𝑜𝑜′ + 𝑐𝑐 [3]

3.4 Plastic Radius (RP)

The Delft equation has been criticized for producing unconservative formation limit pressures in sands under drained
loading conditions (Xia, H., 2009). One parameter that can have a significant influence on the FLP is the plastic radius.
The original Delft research suggested using a plastic radius of two-thirds the depth of the profile in sand soils and one-
half the depth of the profile in clay soils as a means of applying a factor of safety to the results. Staheli pointed out in
her paper (Staheli et. al, 2010) that at RP values greater than about 10 feet there is no appreciable increase in the FLP
and that at values less than 10 feet, the FLP begins to decline rapidly such that at deeper depths (greater than 10 feet),
the recommended values of RP do not significantly impact the FLP and therefore should not be considered as a method
to apply a factor of safety to the results.

Recent discussions with the US Army Corps of Engineers (USACE) Risk Management Center (RMC) have suggested
that limiting the Rp value to something on the order of two- or three-hole diameters might be a reasonable way to limit
plastic deformation of the formation surrounding the hole when completing HDD installations under critical facilities
such as levees.

3.5 Shear Modulus

The deformation properties of elastic materials are described by modulus and Poisson’s ratio (υ). Young’s modulus
(E) is determined during uniaxial loading from the initial linear slope of the stress-strain curve during loading. Note,
soil stress-strain response is non-linear and stress-dependent, but over low stress and strain ranges is assumed to act
more or less linear elastically. Shear modulus (G) actually describes the soil “skeletal” response (Kullhawy and
Mayne, 1990); it is the slope of the shear stress – shear strain curve upon loading and is related to Young’s modulus
by the following equation (4):

𝐺𝐺 = 𝐸𝐸/[2(1 + 𝜐𝜐)] [4]

Geotechnical engineers commonly estimate moduli and Poisson’s ratio from soil strength and index testing; we do not
present those correlations here for brevity.

4. RECOMMENDATIONS FOR DESIGN

During the design stage of a project, the geotechnical engineer should carefully consider several factors to better
understand and characterize the subsurface conditions and thereby reduce the risk of hydraulic fracture and IR during
construction.

The designer needs to consider whether the geotechnical conditions warrant modelling the soils under drained or
undrained conditions. For relatively clean (low fines content) granular soils, utilizing drained soils strength parameters
is warranted. We recommend that if using the Delft Equation, undrained soil conditions be utilized for fine-grained
soils (silt and clay) and that it may also be prudent for coarse-grained soils with significant fines content (perhaps
greater than 40 percent). At this point it’s not clear how quickly pore pressures dissipate in granular soils with
significant fines content and that future research may lead to a procedure that allows the use of drained conditions and
a reduced or “effective” friction angle that can be relied upon for predictive purposes.

If the Delft Equation is being used for drained conditions with the Delft recommended Rp value, a factor of safety of
at least 2.0 is recommended when calculating allowable pressures for crossings where IRs would be detrimental to the
project. If the Rp value is limited to two- or three-hole diameters, we suggest a factor of safety of 1.5 could be
appropriate. To calculate allowable pressures for undrained conditions, a factor of safety of 2.0 or greater is

Paper TA-T2-03 - 5
recommended but the project team should realize that it can be difficult or impossible to maintain annular pressures
low enough to maintain a factor of safety of 2.0 in undrained silts and clays.

5. RECOMMENDATIONS FOR CONSTRUCTION

The most effective method to validate the results of a hydraulic fracture analysis is to monitor the downhole annular
pressures during pilot hole operations. Monitoring the annular pressures serves several purposes but, with regard to
hydraulic fracture analyses, the two main advantages are that the recorded pressures can be compared to the estimated
pressures calculated in the analysis and occurrences of inadvertent returns during drilling may be used to validate the
formation limit pressures provided by the analysis. If IRs are occurring at pressures less than the formation limit
pressure, the input parameters used in hydraulic fracture analysis may need to be re-evaluated.

Monitoring should include continuous recording of the annular pressures into a computer-generated file at a time
interval of 5 seconds or less. Real-time visual monitoring of the annular pressure is also useful for observing the
average annular pressure over the length of each joint of drill pipe and also observing short-term pressure spikes that
can initiate hydraulic fracture.

During drilling, if quick pressure spikes are observed and the pressure then drops substantially below the average
pressures observed over the previous few joints of drill pipe, it’s likely that the formation has been fractured, drilling
fluid is being lost to the formation, and the risk of inadvertent returns is increased. At that point it is prudent to stop
the drilling fluid pumps and assess the situation. If the hole behind the BHA has become plugged it may be necessary
to trip out the BHA to a point behind the blockage so that it can be cleared and drilling can resume at normal annular
pressures and fluid returns rates.

It’s important to point out that, although it’s useful to visually monitor the rate of drilling fluid returns, it can be
difficult to visually determine if full drilling fluid returns are being maintained to the returns pit and as such an observer
may not realize that the drilling fluid returns are only partial, and that fluid is being lost downhole.

6. SUMMARY

The Delft method is a proven model, correlating well with actual trenchless project experience in the authors’
experience. The Delft method is useful in assessing formational pressure limits and the risk of inadvertent drilling
fluid returns during Direct Pipe® and HDD construction, provided appropriate soil strength and other parameters are
used. To properly use the model, an understanding of applied geotechnical engineering is necessary. Trenchless
construction imposes a rapid load on soil and the drainage characteristics of the soils should be considered when
assigning strength and other parameters for assessment.

7. REFERENCES

Bourgoyne, A.T., Millheim, K.K., Chenevert, M.E., and Youg Jr., F.S. (1991) “Applied Drilling Engineering,” Society
of Petroleum Engineers, Richardson, TX.

Kulhawy, F.H., Mayne, P.W. (1990) “Manual on Estimated Soil Properties for Foundation Design” Cornell
University, Prepared for Electric Power Research Institute

Robison, J.L., Sparks, A.E. (2015) – Direct Pipe® Levee Crossing Design – Mitigating Hydraulic Fracture Risk,
NASTT No Dig Show Conference Proceedings Paper WM-T4-04.

Staheli, K., Bennet, R.D., O’Donnell, H., Hurley, T. (1998) – Installation of Pipelines Beneath Levees Using
Horizontal Directional Drilling, U.S. Army Corps of Engineers, Report No. CPAR-GL-98, Waterways Experiment
Station, CPAR-98-1.

Paper TA-T2-03 - 6
 Staheli, K., Price C.G., Wetter L. (2010) – Effectiveness of Hydrofracture Prediction for HDD Design, Paper F-1-01,
North American Society for Trenchless Technology (NASTT) No-Dig Show 2010.

Terzhagi, K., Peck, R.B. (1967) – Soil Mechanics in Engineering Practice, John Wiley and Sons, 2 Ed.

Xia, H. (2009) - Investigation of maximum mud pressure within sand and clay during horizontal directional drilling
(Unpublished doctoral dissertation). Queen’s University Kingston.

Paper TA-T2-03 - 7

View publication stats