Club of Rome: Eco-nomics for an Ecological Civilization
Borehole Magnetic Resonance for Aquifer Characterization - Jordan Furnans
1. Annual Fall Conference 2020
Jo
Borehole Magnetic Resonance
An Emerging Technology for Aquifer
Characterization for Water Resource Applications
Texas Case Studies
Jordan Furnans, PhD, PE, PG
Vice President – TX Operations
Jordan.Furnans@lrewater.com
512-736-6485
2. NOMENCLATURE:
• Nuclear Magnetic Resonance (NMR)
• Borehole Magnetic Resonance (BMR)
• Synonymous – Some people don’t like the word “nuclear”
• Magnetic Resonance Imaging (MRI)
3. • What is the AQUIFER YIELD?
• How much MOBILE WATER is available?
• How does the PERMEABILITY vary throughout
the aquifer?
Water Resource Needs:
– Production and Recharge Well Design
– Productivity/Injection Estimates
– Well Siting and Permitting
– Groundwater Model Calibration
Can BMR technology assist water managers in assessing
an aquifer for WATER RESOURCE INVESTIGATIONS?
4. BMR TECHNOLOGY
Permanent magnets in probe
polarize hydrogen in H2O
molecules
H2O
NMR Signal
S
0
∝𝜙 ∝pore size
T
2
NMR signal is generated and detected only
within a thin cylindrical sensitive volume
surrounding the center of the tool
5. What is BMR Logging?
• Characterizes the volume, structure, and type of fluids occupying the small
spaces (pores) within an aquifer unit.
• Important differences from other geophysical logs:
– Instead of lithologic indicators, BMR tells us about the spaces and fluids therein.
– Can provide a large array of information with one log
• Water Content
• Porosity (independent of lithology)
• Pore Size (volume/surface area)
• Pore Size Distribution
• Bound vs. Mobile Water
• Hydraulic Conductivity
9. Run Caliper Log First…
Borehole too large
to yield useful data
10. Borehole Geophysics
“NORMAL” SUITE “BMR” SUITE
• E-logs
• Natural Gamma
• Temperature
• Caliper
• Guard
• Sonic
• Neutron
• Density (gamma-gamma)
• E-logs
• Natural Gamma
• Temperature
• Caliper
• BMR (aka. NMR)
Switch can be made if field conditions are met: borehole < 12” diameter,
not a preponderance of large washouts. [Can be larger depending upon tool.]
12. Texas Case Studies
Deep well
Large Potential
Screening Interval
Low Yield Expected
• Need to maximize yield
• Keep TDS < Target
13. Texas Case Studies
Deep well
Large Potential
Screening Interval
Low Yield Expected
• Need to maximize yield
• Keep TDS < Target
14. Texas Case Studies
Deep well
Large Potential
Screening Interval
Low Yield Expected
• Need to maximize yield
• Keep TDS < Target
TDS Calculated
From Resistivity Logs
And Mud Resistivity
HIGH K
HIGH TDS
HIGH K
Low TDS
Exclude This?
15. Texas Case Studies
Deep well
Large Potential
Screening Intverval
Low Yield Expected
• Need to maximize yield
• Keep TDS < Target
TDS Calculated
From Resistivity Logs
And Mud Resistivity
Excluded Upper
Portion – TDS Reasons
Too short to
Exclude
18. 2016 2011
2002
TRADITIONAL METHOD:
HYDROLOGIC TESTING
A volume of water is
added, & the water-
level change is
recorded
• Falling head or
slug tests in
boreholes
• Aquifer tests only
in completed
wells
19. Slug Test K-values
Slug Test Estimated Productivity:
335 gpm with 100 ft of
Drawdown
Productivity Estimates Using Falling Head Test Data
Zone #
Depth Interval,
feet
Aqtesolv K-value,
ft/d
Applied Zone
Interval, feet
Calculated T-value,
gpd/ft
Estimated SC,
gpm/ft
Estimated capacity at
100 ft drawdown, gpm
Zone 1 1380-1410 0.83 45 279 0.19 19
Zone 2 1320-1350 1.08 67 541 0.36 36
Zone 3 1245-1275 1.08 68 550 0.37 37
Zone 4 1185-1215 1.45 57 617 0.41 41
Zone 5 1130-1160 1.10 68 562 0.37 37
Zone 6 1050-1080 1.55 103 1190 0.79 79
Zone 7 925-955 0.88 195 1289 0.86 86
SWL
Estimated Total Production Rate, gpm 335
where:
Q = gpm;
T = gpd/ft;
s = feet of drawdown
T
1,500
=
(gpm/ft) Q/s
Empirical Equation
T = Kb
Phoenix ASR Well
20. Phoenix ASR Well
Productivity Estimates using NMR Data
Zone #
Depth,
feet
NMR K-
value, ft/d
Applied Zone
Interval, feet
Calculated T-
value, gpd/ft
*Estimated SC,
gpm/ft
Estimated
Capacity at 100 feet
Drawdown, gpm
Zone 1 1412 1.49 25 278 0.19 19
Zone 2 1387 0.33 25 62 0.04 4
Zone 3 1362 0.51 25 95 0.06 6
Zone 4 1337 1.28 25 240 0.16 16
Zone 5 1312 0.88 25 165 0.11 11
Zone 6 1287 0.78 25 145 0.10 10
Zone 7 1262 0.98 25 184 0.12 12
Zone 8 1237 1.34 25 251 0.17 17
Zone 9 1212 1.73 25 323 0.22 22
Zone 10 1187 6.33 25 1185 0.79 79
Zone 11 1162 2.85 25 534 0.36 36
Zone 12 1137 6.15 25 1149 0.77 77
Zone 13 1112 8.07 25 1509 1.01 101
Zone 14 1087 1.60 25 299 0.20 20
Zone 15 1062 6.75 25 1262 0.84 84
Zone 16 1037 2.30 25 431 0.29 29
Zone 17 1012 22.67 25 4239 2.83 283
Zone 18 987 0.92 25 172 0.11 11
Zone 19 962 0.96 25 179 0.12 12
Zone 20 937 0.67 25 125 0.08 8
Zone 21 912 2.05 25 384 0.26 26
Zone 22 887 2.79 25 522 0.35 35
Zone 23 862 2.48 25 463 0.31 31
Zone 24 837 7.21 25 1348 0.90 90
Zone 25 812 7.53 5 282 0.19 19
SWL 807 605
Estimated Total Production Rate, gpm 870
Well Screen Interval
NMR Zone Estimated
Productivity:
870 gpm with 100 ft of
Drawdown
21. Phoenix ASR Well
335 gpm
100 ft of Drawdown
Slug Test Estimated Productivity:
870 gpm
100 ft of Drawdown
NMR Zones Estimated Productivity:
Aquifer Test Productivity:
1,095 gpm
100 ft of Drawdown
23. Significant underestimation of
hydraulic conductivity from slug
tests
NMR fills in gaps in hydraulic
conductivity rather than interpolation
Lack of zone development can impact
zonal slug test results
Projecting underestimated K-values
exacerbates the problem (i.e.,
underestimated productivity)
Evaluation of NMR K values was not
quick nor simple. This would often need
to be done “on the fly” for well design.
Phoenix ASR Well Take Away
24. Production and Exploration Sites
Ave. NMR in monitor well
K = 13 ft/day 280 – 500 ft bls
K = 4.9 ft/day 500 – 750 ft bls
NMR conducted after
monitor well was
constructed inside 5”
PVC casing & screen
Aquifer test K in
shallow well
= 16.6 ft/day
Aquifer test K in
deeper well
K = 2.6 – 4.7 ft/day
Metro Water District’s
OJ2 Replacement Well
in Tucson, AZ
25. Hydraulic Conductivity
– Aquifer Testing
Aquifer test
K = 60.7 ft/d
BMR Log
Geo. Mean: KSDR = 3.3 ft/d
Geo. Mean: KTIM = 11.5 ft/d
BMR Log
Geo. Mean: KSDR = 1.6 ft/d
Geo. Mean: KTIM = 3.0 ft/d
Aquifer test
K = 28.4 ft/d
Aquifer test
K = 94 –106 ft/d
26. Specific Yield
Aquifer test
Sy = 0.14
BMR Log
Arith. Mean: Sy = 0.12
Geo. Mean: Sy = 0.11
BMR Log
Arith. Mean: Sy = 0.13
Geo. Mean: Sy = 0.10
Aquifer test
Sy = 0.12
Aquifer test
Sy = 0.11
27. Benefits
Direct detection and quantification of water
content and porosity
Determination of bound versus mobile water
Quantitative estimates of hydraulic
conductivity, transmissivity, specific yield, and
pore-size distribution
Sy is a difficult parameter to achieve, and NMR
seems to be a good source.
Challenges
o Confidence of dataset, especially for K
o Borehole diameter limitations
o Requires pre-project planning to ensure
limitations are understood, adequate data is
collected, proper constants are applied for
data processing
o K estimates will get better if it’s used more in
a geographic area
Conclusions
28. Jordan Furnans, PhD, PE, PG
Vice President – TX Operations
Jordan.Furnans@LREwater.com
512-736-6485
Editor's Notes
Gary
Gary – There’s a nomenclature problem associated with this type of logging that we need to acknowledge. When I first commissioned this log, it was called “nuclear magnetic resonance” or NMR. Some stakeholders out there didn’t like the word “nuclear”, so some companies switched to calling it “borehole magnetic resonance” or BMR. There is absolutely no difference between the terms, and because they’re both used, we’re not going to try settle the matter. This talk uses both terms so you can get used to using both terms. By the way, if you’ve ever had the joy of being inside one of those noisy MRI machines, you have experienced the technology firsthand.
Hand it to Lauren
LAUREN
LAUREN
Lauren -
In summary, and in contrast to other types of borehole geophysical logs, BMR is all about the porosity.
Then hand back to Gary
Gary --
This is a log from an exploration borehole in southern Arizona. Using the principles that Lauren shared with us, the BMR system calculates pore size distributions every 0.2 ft. Based on the relaxation times, we see that some of the water in the pores is bound in clays and some is capillary-bound. Clay-bound and capillary-bound water aren't useful for water resource interests. Only the moveable water would be considered a resource.
Gary
The calculated porosity distributions can also lead to hydraulic conductivity. It’s important to note that the two K algorithms are calculations based on the calculated porosity calculations. The other very important point is that the original algorithms are based on consolidated sediments and should be adjusted for local conditions. Thus, the more experience in an area, the better the adjustment.
Gary --
So, if you thought that when I mentioned “movable water” that that sounds a lot like specific yield, you’re right. By definition, Sy is movable water.
Gary --
Lauren mentioned earlier the investigation radius of NMR. Simply, if the borehole is too large, the data are no good. We suggest you run the caliper log first to determine if enough of the borehole will yield useful BMR data. In this case, only one zone, well above the aquifer, was "bad".
Gary –
If you decide to use BMR, you’re not just adding another log to your usual suite - BMR can replace some of the logs in the normal suite. The way I do it is to have the logger bring all their tools, run the caliper log first, and then make a decision in the field whether to run the “BMR” suite or, if the borehole isn’t suitable, then run your “normal” suite.
Hand back to Lauren to talk about testing with hydraulic parameters
Comparative data, examples of reliability/repeatability
Build our confidence level in the technology
Comparative data, examples of reliability/repeatability
Build our confidence level in the technology
Comparative data, examples of reliability/repeatability
Build our confidence level in the technology
Comparative data, examples of reliability/repeatability
Build our confidence level in the technology
Comparative data, examples of reliability/repeatability
Build our confidence level in the technology
Comparative data, examples of reliability/repeatability
Build our confidence level in the technology
Lauren – Phoenix Case Study intro
Lauren
Intro two types of typical testing
Transition into the case studies where we did both traditional and BMR testing, and compare results
Lauren
Lauren – remove animations
Lauren –
hand back to Gary
Gary – Metro Water District had an old, shallow well, and since the regional groundwater level was declining, they needed to replace the well much deeper. The original well is represented by the little green square on this figure. Due-diligence investigation of the deeper aquifer included an exploration boring represented by the small blue square. The large blue circle represents the replacement well. The permeability estimate in the shallower portion of the NMR log compared favorably to the shallow aquifer test, and similarly the K estimate in the deeper portion of the NMR log was very close to the deeper well’s aquifer test results.
Gary-
BMR logging was conducted in two exploratory boreholes (the green on this figure) that were drilled as part of a due-diligence investigation for a new wellfield NW of Tucson (the blue on this figure). In this case, BMR did not compare well to aquifer test results at the recovery wells. I noticed a footnote on the BMR log stating that the coefficients used to calculate the conductivity estimates were bases on consolidated formations. This is an unconsolidated aquifer. We worked with the logging company to refine their coefficients, and the estimates got a little better. I think this illustrates what Parsekian et. al. say in their 2015 Groundwater article – that the more BMR is used in an area, the more “calibrated” the method becomes.
Gary –
On the other hand, Sy from BMR compared amazingly well Sy estimates from aquifer testing as you can see here.
Hand off to Lauren to wrap it up…
Presenter?
UPDATE SLIDE – LAUREN and GARY
Since Gary started, Lauren should wrap it up