EPA/600/R-06/122
November 2006
Development and Demonstration of a
Bidirectional Advective Flux Meter for
Sediment-Water Interface
by
Bob K. Lien
Land Remediation and Pollution Control Division
National Risk Management Research Laboratory
Cincinnati, Ohio 45268
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
Notice
The U. S. Environmental Protection Agency through its Office of Research and
Development funded the research described here. This report has been subjected to
Agency's peer review and has been approved for publication as an EPA document.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
11
-------
Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with
protecting the Nation's land, air, and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and implement actions leading to a
compatible balance between human activities and the ability of natural systems to support
and nurture life. To meet this mandate, EPA's research program is providing data and
technical support for solving environmental problems today and building a science
knowledge base necessary to manage our ecological resources wisely, understand how
pollutants affect our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's
center for investigation of technological and management approaches for preventing and
reducing risks from pollution that threaten human health and the environment. The focus
of the Laboratory's research program is on methods and their cost-effectiveness for
prevention and control of pollution to air, land, water, and subsurface resources;
protection of water quality in public water systems; remediation of contaminated sites,
sediments and ground water; prevention and control of indoor air pollution; and
restoration of ecosystems. NRMRL collaborates with both public and private sector
partners to foster technologies that reduce the cost of compliance and to anticipate
emerging problems. NRMRL's research provides solutions to environmental problems
by: developing and promoting technologies that protect and improve the environment;
advancing scientific and engineering information to support regulatory and policy
decisions; and providing the technical support and information transfer to ensure
implementation of environmental regulations and strategies at the national, state, and
community levels.
This publication has been produced as part of the Laboratory's strategic long-term
research plan. It is published and made available by EPA's Office of Research and
Development to assist the user community and to link researchers with their clients.
Sally Gutierrez , Director
National Risk Management Research Laboratory
in
-------
Abstract
A bidirectional advective flux meter for measuring water transport across the
sediment-water interface has been successfully developed and field tested. The flow
sensor employs a heat-pulse technique combine with a flow collection funnel for the flow
measurement. Because the direction of flow was initially unknown, the heater was
located in the center of the flow tube. Two thermocouples were symmetrically placed to
both sides of the heater for temperature monitoring. For each measurement cycle, the
heater generates and injects a heat-pulse to the center of the flow tube, the water flow
inside the flow tube carries the heat-pulse down gradient, and the temperature is
monitored at each thermocouple over time. In theory, the heat-pulse arrival time is
inversely proportional to the flow rate. The bidirectional feature of the flux meter is
realized through the temperature measuring capability on either side of the flow tube.
The system has automatic data acquisition, real time data display, and signal
display/analysis capability. The instrument has undergone several calibrations to establish
empirical relations between flow rate and heat-pulse travel time. Flow rate can be derived
as a function of peak temperature arrival time, or as a function of first temporal moment
of the heat-pulse. In the field operation, the flow across sediment-water interface is
funneled through a dome to the flow tube; the rate of water flowing through the flow tube
is measured. The advective flux through the sediment-water interface, in term of vertical
Darcy velocity, is calculated by dividing the flow rate by the dome area. The dome
serves as an amplifier to bring the generally low Darcy flow within a measurable range
by the flow sensor. The larger the dome area, the smaller the flow it can detect.
The flux meter has undergone field tests at three very different settings. The first
test site was a shallow turbulent stream at Santo Domingo, Nicaragua. The large
magnitude and frequency of shifts between base-flow and storm-flow caused by rainfall
events during our field deployment prevented us from obtaining reproducible data.
Nevertheless, invaluable lessons were learned. The second test site was at Grand
Calumet River, Hammond, IN, which is a slow moving river with fine texture organic
rich sediment. And the final test site was at a large reservoir with deeper sediment at
Lake Hartwell, SC. The last two field deployments were successful in that the flux meter
was operational, reasonable characterization of flux were obtained, and bidirectional flow
measurement capability was demonstrated.
IV
-------
Table of Contents
Notice ii
Foreword iii
Abstract iv
Introduction 1
Objectives 3
Flux Meter Development 3
1) Hardware 4
2) Software 8
3) Calibration 9
Potential Sources of Error 14
Field Applications 19
1) Santo Domingo, Nicaragua 19
2) Hammond, Indiana 20
3) Lake Hartwell, South Carolina 25
Lessons Learned and Ideas for Improvement 28
Conclusions 29
Acknowledgement 29
References 30
Appendix A Calibration Curve for SN152 33
Appendix B Calibration Curve for SN151 34
Appendix C Computer Code 35
Appendix D Sample raw data file 63
Appendix E Results from Hammond, IN 64
Appendix F Results from Lake Hartwell, SC 68
-------
List of Figures
Figure 1. The basic unit of a proto type advective flux meter 4
Figure 2. Schematic of field implementation of the flux meter 5
Figure 3. Conceptual Drawing of the flow sensor 6
Figure 4. Flow sensor circuit board 6
Figure 5. Thermal noise during near isothermal conditions 7
Figure 6. Software: Data Acquisition Module 8
Figure 7. Software: Signal Display Module 9
Figure 8. Calibration curves for flow sensor SN151 based on the method of
moment 11
Figure 9. Experimental thermograph ploted on log scale. Time zero is the
beginning of heating 12
Figure 10. Illustration of abnormalities in thermographs 12
Figure 11. Calibration curves for flow sensor SN151 based on peak temperature
arrival time 13
Figure 12. Tire inner tube placed around dome cylinder to improve seal to
formation. Photo by Goran Bengtsson, Santo Domingo, Nicaragua 14
Figure 13. Bernoulli Effect is anticipated under the turbulent flow conditions in the
rapid stream. The schematic illustration shows installation of
potentiomanometer for head differential measurement. Photo by Goran
Bengtsson, Santo Domingo, Nicaragua 15
Figure 14. Open cylinder used for advective flow measurement. Photo by Goran
Bengtsson, Santo Domingo, Nicaragua 16
Figure 15. Abnormal thermographs observed in laboratory when a stream of gas
bubbles was introduced through the flow tube and significant ambient
temperature drift was present 17
Figure 16. Apparent wave action causing water flow periodically reversing in
direction (observed at Lake Hartwell, SC), such that rise and fall of the
thermographs are coincidently reversed on opposite side of the heater 18
Figure 17. Illustration of a sudden flow reversal observed in otherwise normal
thermographs. Flow reversal is believed to be due to wave-induced
flow caused by boat activities (observed at Lake Hartwell, SC) 18
Figure 18. Thermographs illustrating upward baseline drift during measurement.
Note that the reference temperature (magenta colored points) is not
showing trend of drifting and is not considered as the cause of the
baseline drift 19
Figure 19. Grand Calumet River section monitored. Photo by Dr. Jafvert, Purdue
University 21
Figure 20. Example of passive method signal response from RC2S. The green
curve indicates dispersion of heat up-gradient in low flow condition 22
Figure 21. Example of active method (5 ml/min addition) signal response from
RC2S showing much distinguishable thermograph. It provides better
data quality and also minimizes heat dispersion up-gradient 22
VI
-------
Figure 22. Temporal flux at location RC2S estimated using both methods of
analysis. Note: negative flux value indicates groundwater discharge to
surface water 24
Figure 23. Lake Hartwell Transect locations 25
Figure 24. Temporal flux at location OB estimated based on peak temperature
arrival. Note: positive flux value indicates groundwater recharge from
surface water 27
Figure 25. Temporal flux at location OB estimated based on normalized first
moments. Note: positive flux value indicates groundwater recharge from
surface water 27
vn
-------
Introduction
Understanding chemical inputs into ecosystems and its availability for biological
uptake is a key factor in estimating the risk potential that a chemical poses to human and
ecological receptors. The technology for assessing risk potential is an ongoing topic of
investigation (U.S. EPA 2006). The traditional approach is to characterize a site by
estimating the location and concentration of chemicals within a volume of interest and
then estimate the ecological risk based on this concentration data. Recently an alternative
option for estimating risk includes flux as a key component in the analysis. Flux is
primarily reported as either volume flux or mass flux. Volume flux (LT"1) is usually
related to the fluid carrying the chemical while mass flux (mass L"2 T"1) is used to refer to
the transported chemical. The mass flux is the flow weighted concentration times the
volume flux. The hypothesis is that ecological systems are in fact impacted by the
integrated mass flux of chemical introduced into the system. Research conducted by
(Bockelmann, Ptak et al. 2001; Annable, Hatfield et al. 2005; de Jonge and Rothenberg
2005) provide the design of flux measurement approaches and successful field
application to quantify contaminant mass fluxes and assess its related ecological risk.
Most efforts to date have been limited to characterizing contaminant transport from
one place to another within subsurface soil or ground water environments. Little research
has been performed to quantify water (or contaminant) fluxes from ground water or
submerged sediments to surface water (U.S. EPA 2000). Impacts from the discharge of
contaminated ground water on the ground-water/surface-water (GW/SW) transition zone
ecosystem have often been ignored, even though this ecosystem provides important
ecological services and is the most exposed to ground-water contaminants. The GW/SW
transition zones at the interfaces are extremely biologically active and capable of
metabolizing chemicals or excreting chemicals as water passes from one environment
into another. This region can dramatically change the risk of chemicals and needs to be
understood. Numerous well-known methods exist for parameter estimation and process
identification in aquifers and surface waters. The transition zone, however, has only in
recent years become a subject of major research interest, and the need has evolved
for appropriate methods applicable in this zone (Kalbus, Reinstorf et al. 2006). Providing
a tool that can measure both the volumetric flux of water and the mass flux of the
chemicals associated with the water would increase our understanding of the active
processes. An example is the influence of caps on contaminated sediments. Much of the
design of a cap is based on the slow diffusion of chemicals through a cap. When
diffusion is the dominant transport mechanism a cap can provide isolation from the
contamination for significant periods of time. However, a small amount of advective flux
will significantly reduce the life expectancy of this engineering system to reduce risk.
Since advective flux is a key component affecting chemical inputs from sediments to
impacted ecosystems (Simmons 1992; Harris 1995; Huettel, Ziebis et al. 1998; Shaw,
Moore et al. 1998; Li, Barry et al. 1999; Jahnke, Nelson et al. 2000; Uchiyama, Nadaoka
et al. 2000; Liu, Jay et al. 2001; Yang, Hwang et al. 2002; Jahnke, Alexander et al. 2003;
Slomp and Van Cappellen 2004), it is important to develop such flux measuring
capability to aid in better environmental risk assessment and management. The direction
of this research is to develop tools that can be utilized to measure volume flux through
-------
the ground-water/surface-water interface, hydraulic properties of the hyporheic zone, and
chemical fluxes.
The field effort to measuring water flux in lake was initially conducted by (Lee
1977) using seepage flux measurement with seepage meters coupled with measurement
of hydraulic gradient through the sediment. This was performed as a part of a water
quality study to evaluate the input of N and P to the lake through groundwater from a
nearby septic tank. Recharge to a body of surface water occurs when the hydraulic
gradient at any recharge point in a flow system exceeds the downward-directed gravity
force. Conventional design for measuring in situ (seepage) water flux cross sediments
consists of a submerged chamber (e. g., an inverted dome) connected to a piece of tubing
and a thin plastic film collection bag. Water flux released through sediment-water
interface is calculated by measuring the change in weight of the bag after some time has
elapsed (Lee and Cherry 1978). Water flux measurement made with this method,
however, can be highly variable. Concerns have been raised about the accuracy of these
devices (Fellows and Brezonik 1980; Shaw and Prepas 1989; Shaw and Prepas 1990;
Belanger and Montgomery 1992; Shinn, Reich et al. 2002). The published literature
provides examples of the transient effects of meter installation on flows through the
lakebed (Shaw and Prepas 1989), head loss related to the ratio of meter diameter to
tubing diameter (Fellows and Brezonik 1980; Belanger and Montgomery 1992), and
errors associated with bag type and size (Fellows and Brezonik 1980; Shaw and Prepas
1990; Schincariol and McNeil 2002). Nevertheless, numerous field and laboratory studies
have demonstrated these devices can yield accurate data provided adequate care is taken
in making measurements and proper calibration coefficients are applied.
The original method was modified by (Murdoch and Kelly 2003) by adopting
equilibrium pumping out from the submerged chamber. The nearby axial piezometer was
also used to determine the vertical hydraulic head gradient through the sediments
encompassed by the submerged chamber. The established constant equilibrium pumping
rate (L3 /T) without head change divided by cross section (L2) area of submerged
chamber is equal to seepage rate (L/T). However, this flux measuring method has a
limited application only to shallow depth with relatively high vertical hydraulic
conductivity (-108 m/day). Several researches have modified the original approach of
Lee (1977) by replacing the sampling bag with a flow meter which permits near
continuous measurements. The approaches include heat-pulse methods (Taniguchi and
Fukuo 1993), ultrasonic methods (Paulsen, Smith et al. 2001), electromagnetic flow
meters (Rosenberry and Morin 2004), and dye dilution approaches (Krupa, Belanger et
al. 1998; Sholkovitz, Herbold et al. 2003).
Data from numerous investigations where seepage meters were used indicate flux
through the lakebed commonly varies over three orders of magnitude and can vary by as
much as five orders of magnitude. Results also have shown that spatial variability in
seepage can be considerable, even on a scale of only a few meters (Shaw and Prepas
1990). Seepage measurements have been shown to correlate well with oceanic tides (Lee
1977; Sebestyen and Schneider 2001), typically varying inversely with tidal stage by as
much as an order of magnitude in areas where ground water discharges to the ocean or
estuaries. Seepage rates also have been related to lake seiche (Taniguchi and Fukuo
1996). Modeling investigations have also been conducted studying the importance of
wave and tidal actions showing significant flow reversals ranging in time from diurnal
-------
tidal reversals to a few seconds to tens of seconds for wave action (e.g.(Li and Barry
2000).
Considering the extent of field investigations dealing with seepage processes,
relatively little effort has been directed toward examining the temporal variability of
seepage flux (Sebestyen and Schneider 2001). Originally this is due primarily to the
fundamental limitations of the half-barrel, bag type measurement scheme. If fluxes are
very small, it can take from hours to days to accumulate enough volumetric flow to make
a single reliable measurement. Moreover, each measurement is time integrated, and no
information related to temporal variations within the measurement span is available.
Several automated seepage meters have been developed to monitor transient, as well as
longer-term, seepage effects through sediments in rivers, lakes, and oceans (Taniguchi,
Burnett et al. 2003). Currently, none of these devices is available commercially.
The initial phase of the work has been to develop a robust methodology to
quantify the advective volumetric flux through a plane normal to the instrument. In
addition methods have been demonstrated to use this information to calculate the
hydraulic properties of the formation materials. Methodology to measure the chemical
flux remains to be accomplished and was not included.
Objectives
There is a significant amount of interest in the movement of water between
surface water bodies and ground water. The interest ranges from ecological studies
concerned about turnover rates of chemicals in benthic environments, water resources
studies evaluating gain and loss of water from surface water bodies, to contamination
studies involved in containment transfer or remediation of contaminated sediments. The
original impetus for the research was an interest in determining the long term viability of
sediment caps under conditions of advective flux. In a no flow condition caps should
provide an effective low cost method of isolating contaminants since diffusion is a very
slow process and a cap of as little as a meter in thickness could isolate the contamination
potentially for decades. In this project, the objectives are to develop a method of
measuring the bidirectional advective flux of water at the sediment-water interface;
demonstrate the viability of the flow sensor in a laboratory environment and; combine the
flow sensor with a flow collection funnel that can be implemented in a variety of field
settings. The flow range of interest was slightly less than 1 cm/day to as much as 20
cm/day. For much of the work, only a single flow collection design was used. However,
when a different flow range is expected this could be modified by changing the diameter
of the collection funnel. This report covers the design and operation of the instrument,
calibration of the flow sensor and measurements made at selected field locations.
Flux Meter Development
The advancement of flux meter involved development of the hardware and
software, as well as laboratory testing and calibration.
-------
1) Hardware
With the objectives in mind, a prototype advective flux meter capable of measuring
bidirectional water flow across sediment/water interface has been developed (Figure 1).
The operational principle of this device is based on the concept of heat-pulse technique,
which has been widely utilized as a proven method for a variety of measurements
(Campbell, Calissendorff et al. 1991; Choi, Lee et al. 1991; Taniguchi and Fukuo 1993;
Paillet, Crowder et al. 1996; Sonnenschmidt and Vanselow 1996; Bilskie, Horton et al.
1998; Ren, Kluitenberg et al. 2000; Basinger, Kluitenberg et al. 2003; Heitman, Basinger
et al. 2003; Mortensen, Hopmans et al. 2006). The basic unit consists of a dome (flow
collecting funnel), a flow tube, a USB interface box, and a laptop computer. The dome
unit consists of a stainless steel dome and a polyethylene cylinder joined and sealed by
gasket material and secured with four cables. This two-piece dome design allows
intermittent measurements to be made at the same location by leaving the cylinder in the
sediment between measurements. The dome and the flow tube are connected, and the
water flows in or out of the unit dependent on the direction of the gradient. The
electronics of the flow tube are connected through a cable to the USB interface
box/circuit board which provides power to the flow tube. A USB cable connects the
interface box to a laptop computer which contains the operational software. A vent pipe
is extended from the top of the dome to the atmosphere to reduce the interference of
sporadic gas flow, which can be a significant problem when the bottom sediment is
biologically active.
For field implementation as illustrated in Figure 2, the cylinder portion of the dome
unit is pushed into the sediment to isolate the flow path so that only water flow vertically
through the isolated section of the sediment should flow through the flow tube. The
electrical cable connects the flow tube to the data acquisition unit on the shore or on a
boat. In field operation, the unit is powered by 12V batteries and inverters.
Figure 1. The basic unit of a proto type advective flux meter.
-------
Field Implementation
PC running Control and
Data Acquisition Software
X
F\ux Meter USB interface
and circuit board to supply power to heater
12-wire Cable (120 ft)
Gas vent
(2) batteries (each 12-V, 55 Amp-hr)
for running PC and for supplying
power to heater.
Interfacial Area (0.234 m2
Figure 2. Schematic of field implementation of the f\u\ meter.
The key component of the fiux meter is the flow sensor which is schematically
illustrated in Figure 3. It consists of a flow pipe, a heater, type T thermocouples,
instrumentation amplifiers and reference temperature junction. The entire flow sensor
unit was encapsulated inside a plexiglass outer tube with resin for electrical insulation
and water seal. Because the direction of flow was initially unknown, the heater was
located in the center of the flow pipe. There is one thermocouple at the heater to measure
the heat-pulse that's been generated. Two thermocouples were symmetrically placed to
both sides of the heater for monitoring temperature. All thermocouples are connected to
the same reference temperature. The thermocouple reference junctions are maintained at
a common temperature and this temperature is measured using the forward drop across a
semiconductor. Microvolt level signals are amplified at the flow sensor using micro
powered instrumentation amplifiers capable of producing output indicating differential
temperatures with a sensitivity of approximately 100 mV/°C. The absolute temperature
of the reference junction is measured as mentioned above using the forward voltage drop
across a P-N junction with an output of approximately 10 mv/°C. The reference
temperature is calibrated over a range of temperature of 0 - 50 °C with an accuracy of ±2
°C. The thermocouples are used only for relative temperature and are not calibrated. The
gain of the instrumentation amplifier was selected to yield approximately 100 mV/°C.
The nearest standard resistor was used in the circuit. The flow sensor circuit board layout
is shown in Figure 4.
-------
0.1L
SS Resistance heater
51 mm long 30 ohms
ChO
Jacketed Copper
Constantan Thermocouple (T)
Reference Temperature
measured using PN junction
L = 70 mm
Figure 3. Conceptual Drawing of the flow sensor.
FlowMeter REV 1,0 9/7/05 CGEnfield <
Figure 4. Flow sensor circuit board.
For each flow measuring cycle, the heater generates and injects a heat-pulse to the
center of the flow tube, the water flow inside the flow pipe carries the heat-pulse down
gradient, and the temperature is monitored at each thermocouple over time. In theory, the
heat-pulse arrival time is inversely proportional to the flow rate. Therefore, the higher
the flow rate the shorter time it takes for the heat-pulse to arrive. The bidirectional
feature of the flux meter is evident by the heat-pulse measuring capability on either side
of the flow tube. In the field deployment, the rate of water flowing through the flow tube
is measured. The advective flux at the sediment-water interface, in term of vertical Darcy
velocity, is calculated by dividing the measured flow rate to the dome area. While Darcy
velocity may range several orders of magnitude, the detection range of the flow sensor is
limited to a range of 1 to 40 ml/min. Dependent upon the specific study, the size of the
dome may require changes to accommodate the Darcy flux range. The dome serves as an
amplifier to bring the generally low Darcy flow to a measurable range by the flow sensor.
The larger the dome area, the smaller the flow it can detect.
The heat-pulse flow measurement approach has a set of tradeoffs related to heat
losses, noise that is reflected in error of temperature measurement and precision in
-------
determining how fast the heat-pulse is traveling. The advective flux is interpreted from
an analysis of time and temperature data collected at each of the thermocouples. Each of
the temperature data points reported is an average of N observations. The data collection
algorithm could be modified to average out the observed noise by increasing the number
of observations at the expense of increasing the time between sampling events. A
decision was made to use N = 100 as a reasonable compromise. Laboratory analysis of
the measurement system indicates the noise on any given data channel is approximately
±0.02 °C. The laboratory data shown in Figure 5 indicate that under near isothermal
conditions (variation of approximately 0.03 °C (12.93-12.96 °C)) over a period of 12,000
seconds the time weighted average differential temperature did not change significantly
but the noise in the differential temperature measurement is approximately ±0.02 °C. If
the temperature of the measuring thermocouple approaches the temperature of the
reference junction, the differential temperature measurement will simply reflect the
random noise. Thus, the smaller the magnitude of the thermal peak the greater the
potential error in determining the heat-pulse arrival. When flow is low, much of the
energy is lost to the surrounding environment. For this reason, one pair of thermocouples
is placed close to the heat sensor and is the preferred detector under low flow conditions.
The primary differences between the flux meter developed in this project and the
flow sensor developed by (Taniguchi and Fukuo 1993; Taniguchi and Fukuo 1996) are
the bidirectional measurement capability of ours, the type of temperature detector
(thermocouples vs. thermistors) and methods of data interpretation (using both moment
analysis and peak temperature arrival time vs. just peak arrival).
I --0.01
.13
4000 6000
8000
12000
ET(sec)
Figure 5. Thermal noise during near isothermal <
10000
conditions.
•0.15
-------
2) Software
The key controlling mechanism of the flux meter is the software. The code is
written in Visual Basic 6.0 (see Appendix C). The data acquisition module (Figure 6)
provides operational control and data acquisition. It allows changes to various operation
parameters such as the site description, duration of baseline readings, duration of heating
cycle, duration of data collection, number of test cycles, pause-time between cycles, etc.
It displays raw data of each channel in real-time during each test cycle. For each round
of test cycle, a raw data file is automatically generated and named after the beginning
date & time of the cycle. Data is stored in ASCII format (see Appendix D) on the
computer's hard drive with the extension (.dat). The data is kept in its raw form and only
evaluated within the analysis algorithm. The first line of the data file contains
information about how the test was performed, the flow sensor that was used, the
diameter of the dome and the sampling location. From the second line on, the
accumulative time and thermocouple responses from all 5 channels are recorded until end
of the test cycle.
fr Main - [Scan Analog Input]
. Fife Model Tools Window Help
BBS
SileDesciiption [Lake Hjilwei
7/7/2006 9:56:55 AM
Colled baseline signals lot pjQseconds
Stall healer and collect signals lot m seconds
Continue to collect signals loi 11800 seconds altei healei is oil
Pause 1530 seconds beloie next sampling lound
Dome Diameter = [55
Sensoi Seiial Numbei 152
Differential Temperature C
Total sampling founds - pjrj
Curtent sampling lound = 1
Channel 0:
Cable end ° om
Channel 1: 4.030
Healei [2): 2.25G
Channel 3:
Channel 4:
Outlet end
net Temp. (5):
-0.138
-0.211
19.78
Figure 6. Software: Data Acquisition Module.
The signal display module (Figure 7) allows one to choose which data file to
display and process. It automatically finds the calibration for the flow sensor used. It
then analyzes the data and displays the thermographs and the results. The data display
shows relative temperature of the individual thermocouples on the "yl" axis and absolute
temperature of the reference junction on the "y2" axis and elapsed time on the x axis
where t = 0 is the beginning of the heating cycle. The relative temperature is the
difference in temperature between the reference junction and the temperature in the flow
-------
tube at the measuring thermocouple. The analyzed results are stored in a separate file
with the same name as the raw data file except with the extension (.csv) which is directly
readable by Excel. Two primary approaches are used for data analysis. The first is
related to the arrival time of the peak temperature detected at each thermocouple and the
other is the normalized first moment of the thermal breakthrough curve.
Channel 0
Channel 1
heater
Channel 3
Channel 4
Reference
Temperature
Figure 7. Software: Signal Display Module.
3) Calibration
The calibration process involves passing water through the flow sensor at known
flow rates using a precision metering pump. Temperature response is monitored at each
thermocouple and the reference temperature junction as a function of time. Prior to the
injection of heat-pulse a baseline temperature is recorded for a known period of time
(adjustable under program control). Heat-pulse is then injected to the center of the flow
tube. A 10 second heating cycle has been experimentally determined as a desirable
heating time and used as a default value. Although the duration of heating cycle is
adjustable under program control, this value should not be changed unless calibration is
available for that particular heating cycle duration. It is of importance to recognize
calibration curves are dependent on the duration of the heating cycle. Once the duration
is determined it should be maintained throughout the calibration process. Finally,
temperature is recorded during the heating cycle and continued for a period of time after
heating. There are 6N discrete temperature measurements for each test cycle. N is the
number of time steps in a complete cycle. The number of time steps is dependent on the
speed of the computer used to collect the data as well as the speed of the interface card's
-------
analog to digital converter. The system is operated at the maximum speed possible to
minimize magnitude of the time step. The error in time measurement is then dependent
on the sampling frequency which is computer dependent. The data analysis package
displays all of the data graphically. Occasionally random noise is observed in the
response. These noisy data points are obvious to a casual observer because they do not
follow the expected temporal decay pattern in the data and can be eliminated without
statistical analysis. The analysis algorithm currently does not reject abnormal data. This
must be performed manually by the user.
Initially a base line analysis is performed. Both the mean value and slope of the
baselines are determined. The drift in reference temperature is displayed on screen. If the
user has selected to correct for baseline drift, all of the raw data may be adjusted for the
drift in the base line prior to determining peak arrival time or moment analysis. The
duration of the data series should be at least two times the time reported for the first
moment. The baseline is assumed to be a linear function based on the temperatures
measured before the heat-pulse and the last 30 data points from the data set. Baseline
offset is adjusted prior to data display for either approach. Peak arrival time and
amplitude is determined by sequentially searching each data record for the maximum
observed voltage. Note that if there was a spike in the data an incorrect value may be
determined. The user must look at the selected value to make sure that it is reasonable.
Moment analysis is performed on the same data to determine the time of travel for the
center of the heat mass. To include a significant portion of the response curve, the
experimental data collected after 80% of the peak arrival time is fit to an exponential
curve. This removes the potential for negative values and is the approach commonly
used in other tracer experimental analysis. This minimizes the difficulty in handling
values as they approach zero. Standard algorithms are used for calculating the
normalized first moment (Shook and Forsmann 2005; Wu 2006). Computer code for the
first moment calculation is included in Appendix C.
An empirical function which describes the flow of heat though the flow sensor is fit
to the calibration data. Heat flow is controlled by two primary components. At low flow
rates the predominant component is heat conduction through the media. At high flow
rates the primary component is advective flow of the water. The equation describing heat
flow through the transducer based on normalized first temporal moments is:
1.1
M + c2
Where
Q is the volumetric flow rate (L3Tl)
M is the normalized first temporal moment (T)
ci, C2 and C3 are constants
10
-------
E
1
I
a:
I
1451-0,02557
7-6420
e-
829.9 + 0.042T
. ChO
• Ch1
Ch3
Ch4
^Ch 0 Func
— Ch 1 Func
Ch 3 Func
Ch 4 Func
0 =
-1756 + 0.5817-
-2835+06097
T + 5702
0 200 400 600 800 1000 1200
First Temporal Moment (sec)
Figure 8. Calibration curves for flow sensor SN151 based on the method of moment.
An example calibration curve is shown in Figure 8. Data points in the figure
represent the average of N replicate measurements at the same flow rate with N typically
equal to ten. For the purpose of distinguishing flow direction in the flow tube, flow
vector is assigned. Positive flow is considered as flow toward the cable end of the flow
sensor and negative flow is flow away from the cable end of the flow sensor. Moment
analysis is performed on experimental data after initially measuring the differential
temperature for a period of time to establish a baseline. This takes into consideration
both the differential temperature at time zero as well as instrumentation amplifier offset
voltages. At the end of the baseline measurements the data are averaged and subtracted
from the remaining measurements prior to moment analysis. The starting time for the
moment analysis begins with the first sample after baseline measurements. Data analysis
begins with the first observation after the end of the heat-pulse. To minimize errors in
moment analyses, the thermograph is divided into two parts (Figure 9). The first portion
of data, beginning at the end of the heating cycle and continuing to the time when the
temperature has dropped to 80% of the peak temperature, is analyzed directly without
interpolation and extrapolation. From 80% of peak temperature until the signal is less
than or equal to 30% of the maxim temperature measured, the data is fit to a logarithmic
function. This function is used to interpolate and extrapolate anticipated temperatures in
the moment calculations. It is necessary to fit the tail portion of the breakthrough curve
to a function which permits extrapolating the data to very small values. The noise in the
temperature measurements that occurs during heat-pulse switching is illustrated in Figure
10. We were not able to remove this noise using analog techniques on the circuit board.
11
-------
End of direct moment analysis
d on experimental data
Beginning of data fit to function
used to interpolate/extrapolate
remaining response
Figure 9. Experimental thermograph ploted on log scale. Time zero is the beginning of heating.
Tefliper «(«
1.1
1
OS
08
| 0.7
-------
later the coefficient of variation in repeated measurements of the same flow rate is greater
for the peak arrival time than it is for the time for the normalized first moment. This
suggests that using moment analysis may be a preferred approach for data interpretation.
The empirical equation describing heat flow through the flow sensor based on the
peak temperature arrival time follows:
= ct
Where
Q is the volumetric flow rate (L3Tl)
t is the time when the peak temperature is observed (T)
c and b are constants
An example of calibration curves based on peak temperature arrival is shown in Figure
11.
100 200 300 400 500
Peak Arrival Time (sec)
600
700
800
Figure 11. Calibration curves for flow sensor SN151 based on peak temperature arrival time.
Both peak temperature arrival and method of moment techniques provide good
calibration results as demonstrated in Figure 8 & Figure 11. The coefficient of variation
for repeat measurements under laboratory conditions of near constant temperature and
constant flow using a peristaltic pump are typically 2% based on the method of moments
and 4% based on peak temperature arrival. The R2 values are also consistently better
when the method of moments is used, but both methods typically give R2 of 0.99 or
better. Just because laboratory measurements suggest moment analysis will provide a
better result does not mean the results under field conditions are better using moment
analysis. Variability in flow during the measurement may cause greater errors in the
moment analyzed response curves than in the peak temperature arrival time analysis.
13
-------
Potential Sources of Error
• flow bypassing
• Bernoulli Effect
• dome geometry
• gas flow
• wave action
• thermal drift
The primary sources of environmental measurement error are related to flow
bypassing and unstable flow. If the flow that enters or exits the dome through the
sediment interface and respectively exits or enters at a location other that the sensor, there
will be an error in the measurement. There is potential for leaks at the interface between
the dome and the sediment as well as joints and fittings in the dome. For rocky
formations where it is difficult to obtain a seal, a flange can be installed on the cylinder to
reduce the probability of flow bypassing between the dome and the sediment (as shown
in Figure 12). Sandbags could be placed on the inner tube to help create a seal. To
evaluate these potential errors in the field either a standard addition or standard
subtraction is used. The standard addition method is performed by measuring the flow
rate and then immediately repeating the measurement while adding a known flow rate to
a port on the dome. The resulting measurement should be the sum of the two flows.
Flows may change diurnally but the error associated with measurements closely spaced in
time should be minimal after an initial stabilization period. The flow meter is calibrated
over a finite range of flows. If the existing flow is close to the upper limits of calibration,
it may be preferred to perform a flow subtraction rather than a flow addition.
Conversely, flow addition is the preferred method when flow is near zero. Flow
additions when the flow is near zero may be used to refine the measurement such that a
more sensitive region of the calibration curve is used.
Figure 12. Tire inner tube placed around dome cylinder to improve seal to formation. Photo by
Goran Bengtsson, Santo Domingo, Nicaragua.
14
-------
A second source of potential error is related to the Bernoulli Effect that may cause
the head in the dome to be different than the head outside the dome (Libelo and
Maclntyre 1994; Shinn, Reich et al. 2002). This is possible in rapidly moving streams
but not considered to be a significant source of error in slow moving bodies of water such
as lakes or estuaries. If there is a potential for significant Bernoulli-induced flow, the
pressure difference is measured manually by measuring the head difference using tubing
as a water manometer. When a gradient exists the discharge point of the flux meter is
rotated to find the direction where the gradient is no longer significant. Figure 13 shows
a testing of head differential between inside the dome and the stream at the flow meter's
discharge point. The potentiomanometer used for the direct measurement of differences
in hydraulic head is similar to the design of (Winter, LaBaugh et al. 1988). One of the
tubes is inserted through the vent pipe to inside the dome and the other is placed in the
stream near the discharge point of the flux meter. Hand vacuum pump (far left of photo)
was used to create a common vacuum allowing the measurement of the differential
pressure. The discharge point on the flux meter is adjusted to zero the pressure
differential between inside the dome and outside the dome.
vacuum
potentiomanometer
Figure 13. Bernoulli Effect is anticipated under the turbulent flow conditions in the rapid stream.
The schematic illustration shows installation of potentiomanometer for head differential
measurement. Photo by Goran Bengtsson, Santo Domingo, Nicaragua.
Dome geometry can have a significant impact on errors. Three geometries have
been considered each having potential issues. An open cylinder, as shown in Figure 14,
that has one end penetrating the sediment-water interface has desirable characteristics of
ease of placement, not influenced by gas production in the sediment and low cost. The
potential shortcoming of this method of measurement is when the water table is changing
as would be the case in tidal environments. For example if the water table were to
15
-------
increase 1 mm during a 10 min measurement time and the dome diameter were 50 cm,
the change in water volume within the dome would be almost 200 ml or 20 ml/min. With
this much change during the measurement, it would not be possible to obtain any relevant
data. To correct for this error a pressure transducer would be needed to monitor the head
within the dome during measurements.
Figure 14. Open cylinder used for advective flow measurement. Photo by Goran Bengtsson, Santo
Domingo, Nicaragua.
An alternative approach would be to place a sealed dome over the sediment-water
interface where there would be no change in volume. This would eliminate the problem
of changing hydraulic heads in the water body. However, if there is gas accumulation
inside the dome it would potentially dampen the water flow because of its compressible
nature until sufficient gas accumulates in the dome to cause gas flow through the sensor.
Gas flow would create other problems in flow measurement due the different thermal
properties of the two fluids. The effect of gas bubbles can be illustrated by Figure 15,
which shows abnormal thermographs observed in laboratory when a stream of gas
bubbles was introduced through the flow tube and significant ambient temperature drift
was present during the measurement cycle. Drifting in the reference temperature was
most likely also the cause of the thermal drift in the thermographs.
A third alternative is a compromise between the two approaches. A small diameter
pipe (l/2" NPT) has been selected to provide a vent for escaping gas as shown in Figure 1.
This will minimize the gas problem which will show up as noise in the data rather than
having a consistent pattern. The error in flow will be relatively small under slowly
changing conditions similar to what was seen at Hammond Indiana. The nominal inside
diameter of a Schedule 80 Va" NPT is 1.4 cm. Thus, the volume change within the dome
for a 1 mm change in head would be 0.16 ml which results in a 0.016 ml/min using the
conditions described above. With a target detection limit for flow-rate of 1 ml/min this
would not be a quantifiable error. A pressure transducer would still be necessary when
fluctuations are expected to be rapid or large but may not be necessary under quiescent
16
-------
environmental conditions. In large lakes and estuaries, wave action is likely to be 300
mm or more within a period of less than a minute (Li and Barry 2000). If one assumes
the period is one minute and the change in head is 300 mm, this would be equivalent to
46 ml/min (for the Va" NPT pipe on a 50 cm dome) flow reversal every minute as
illustrated in Figure 16. Note that the green and red curves are the thermocouples on one
side of the heater and the blue and cyan curves are on the other side of the heater. When
the red and green curves show an increase in temperature, the blue and cyan curves
coincide with a decrease in temperature, suggesting water flow periodically reversing in
direction. The period appears to be approximately every two minutes. The phenomena
caused by the apparently wave-driven advective flow (Precht and Huettel 2004) makes it
difficult to correctly determine the peak temperature arrival time and an accurate moment
analysis would also be nearly impossible. Even in locations where the water surface is
relative quiescent, boat traffic may create a significant water surge and sudden flow
reversal as shown in Figure 17. A possible solution to this problem is used of a closed
dome that is only vented periodically to remove gas accumulation in between test cycles.
An approach utilizing a spring loaded port door (Menheer 2004) may be applicable for
this purpose. Note that even such a modified dome may still not be ideal in all
circumstances.
Temperature Response
Channel 0
Channel 1
heater
Channel 3
Channel 4
Reference
Temperature
so
100
200 250 300
Time (sec)
350
500
Figure 15. Abnormal thermographs observed in laboratory when a stream of gas bubbles was
introduced through the flow tube and significant ambient temperature drift was present.
17
-------
Temperature Response
Figure 16. Apparent wave action causing water flow periodically reversing in direction (observed at
Lake Hartwell, SC), such that rise and fall of the thermographs are coincidently reversed on opposite
side of the heater.
Temperature Response
Channel 0
Channel 1
heater
Channel 3
Channel 4
Reference
Temperature
200 300 400
Time (sec)
500 600
Figure 17. Illustration of a sudden flow reversal observed in otherwise normal thermographs. Flow
reversal is believed to be due to wave-induced flow caused by boat activities (observed at Lake
Hartwell, SC).
A final potential source of error can occur when the flux meter is initially placed
into a new location where the temperature of the water body is different than the
reference temperature of the flow sensor. When this happens, a drift in baseline
temperature is observed. This creates errors in determining when the maximum
temperature is observed or in the moment analysis. Corrections can be made in the data
analysis by adjusting the measurements based on the slope of the baseline temperature.
This correction should not be made when the drift is very small; small enough such the
18
-------
reference temperature is capable of tracking the temperature of surrounding media. An
analysis as to when a correction should be made and when it shouldn't be made has not
yet been developed. The rule of thumb is when the drift on the baseline is smaller than or
equal to the noise (0.02 °C), then correction should not be necessary. The data analysis
software permits either considering baseline drift or ignoring baseline drift at user
discretion. Raw data as displayed in Figure 18 can be adjusted under program control
(signal display module) to correct for baseline drift. The assumption that is made is that
the drift is linear. Data collected prior to starting the heater is averaged and the last 30
data points are employed. These two averaged values are used to determine the slope of
the baseline and the thermograph is adjusted to remove baseline drift. The raw data file
is not modified, only the file containing the interpreted data is modified.
Temperature Response
0.32
03
0:28
OJ26
024
021
8 0.2
«0.18
a018
gO.14
0.1
008
OJK
0.04
0.02
0
-0.02
baseline drifting upwarc
50 100 150 200 250 300 350 400 450
26.6
2679
2678
26.77
26.76
2S.7S
26.74
26.732
26.72 £
26.71 «
26.7 |
26 69 «
26.68
26.67
26*6
26.65
26.64
26.63
26.62
Channel 0
Channel 1
heater
Channel 3
Channel 4
Reference
Temperature
Figure 18. Thermographs illustrating upward baseline drift during measurement. Note that the
reference temperature (magenta colored points) is not showing trend of drifting and is not
considered as the cause of the baseline drift.
Field Applications
Field demonstrations of the advective flux meter were conducted at three very
different environmental settings: (1) a fast moving and turbulent shallow stream with
cobbly stream bed where Bernoulli Effect is expected; (2) a slow moving river with fine
sediments where flux is expected to be low; and (3) a large reservoir with deeper
sediments where lake seiche and wave action are expected.
1) Santo Domingo, Nicaragua
This field demonstration was a joint effort with Dr. Goran Bengtsson, University
of Lund Sweden to investigate the magnitude of groundwater discharge in a mercury-
contaminated stream near Santo Domingo, Nicaragua. Santo Domingo is a gold mining
town in the mountains of Nicaragua. There are frequent intense rainfall events with large
19
-------
amounts of runoff causing large swings in the hydrograph. Surface water is relatively
fast moving and turbulent. When water flows over and/or around the flux meter, the
sudden increase in water speed creates a pressure difference between the inside and the
outside of the dome, and thus creates a Bernoulli-induced flow through the flux meter.
Based on our observations, the Bernoulli Effect could cause more than ± 10 cm water
differential pressure depending on how the dome is oriented in the stream. Just a few
degrees of dome rotation could create a significant difference in gradient.
As shown in Figure 13, the flux meter was not fully submerged because of the
shallow water depth. Therefore, an open cylinder was used (Figure 14) instead of the
closed dome. The flow sensor was installed below water and through the wall of the
cylinder. Flow sensor SN152 was used at this site (see Appendix A for the calibration).
The stream's bottom sediment was rocky. The flexible flange made from an inner tube
appeared to provide an adequate seal to the formation. However, due to the large swings
in hydrograph caused by the frequent rainfall events during our field deployment, we
were not able to obtain reproducible data for analysis.
2) Hammond, Indiana
This field deployment is part of a joint study with Dr. Chad Jafvert, Purdue
University to investigate capping as a potential remedy for the polycyclic aromatic
hydrocarbon (PAH) contaminated sediments in Hammond, IN. The Grand Calumet
River is a slow moving river with its source primarily originating at the discharge from a
sewage treatment plant. The river sediment is composed generally of silt-sized particles
with a very high percentage of organic matter, consisting of both natural organic matter
and coal tar as previously reported (Jafvert, Lane et al. 2006). Below this 8-12 ft organic-
rich sediment layer was 1-2 ft of fine to medium grain sand layer over a continuous
impermeable clay layer. The sand layer was extensive enough to be connected, but not
always present.
Purdue University conducted a preliminary testing of the advective flux meter
(flow sensor SN151) from mid July to late October, 2005. Because the hydraulic
conductivity within the sediment of the Grand Calumet River was expected to be low due
to its silty-organic texture, the seepage meter was calibrated for low flows (the calibration
is shown in Appendix B). Note that the flow vector was not designated in this
calibration; flow direction was determined by the orientation of the flow tube and the
channels showing heat-pulse response. In this particular case, discernible response on
channel 3 & 4 indicates groundwater discharge to surface water; response on Channel 0
& 1 points to groundwater recharge. In order to allow for the calculation of sediment
hydraulic conductivity, the piezometric head gradients at each measurement location was
measured. The piezometers and stream gauges were installed manually by pushing to the
target depth. In the river, six piezometer clusters were installed, with each cluster
consisting of two piezometers pushed to depths of 4 and 8 ft below the sediment-water
interface, and one stream gauge, each located approximately 6" apart. The location of
each cluster is shown on Figure 19. River clusters 1 to 4 (RC1 to RC4) were installed in
the center of the river at an interval of approximately 80 ft from upstream (RC1) to
downstream (RC4). River clusters RC2S and RC2N where installed at the same
downstream location as RC2, approximately 3-4 ft from the south and north banks of the
20
-------
river, respectively. Two additional stream gauges, SGI and SG2, were installed
approximately 250 ft upstream of RC1 and 100 ft downstream of RC4 to measure the
total horizontal hydraulic head gradient of the river. Over this period, the average
horizontal hydraulic gradient was 3.1 xiQ"4 (± 0.4 xlO"4). The river elevation steadily
decreased over time from an average depth at each of the river clusters of approximately
3.5 ft in August to approximately 2.5 ft in October.
'•
Figure 19. Grand Calumet River section monitored. Photo by Dr. Jafvert, Purdue University.
At each river cluster, the vertical hydraulic head gradient was calculated by
dividing the hydraulic head differences between the piezometers and stream gauge by the
depth difference. The water levels within the 4 and 8 ft piezometers were generally
higher than the river elevation, with all 3 levels generally decreasing over the 3 months of
continuous measurement from mid July to late October, 2005. The vertical gradients
indicate upward vertical water flow with a gradient of 0.1 to 0.2. The temporal changes
in the gradients were minimal over the measurement period, except after high rainfall
events when the change in elevation of the river was much more significant than the
changes in water levels within the piezometers. This probably results from the difference
in the characteristic response times of the piezometers compared to the stream gauges.
The flow of water across the sediment-water interface was measured with the flux
meter within 1-2 ft of each piezometer cluster on at least two different days at each
location. At each location, the cylinder was pushed into the sediment and the dome was
secured with the four cables. After all flow measurements, the dome was removed
leaving the cylinder in the sediment, which allows for returning later to make
measurements at the same location by re-attaching the dome to the cylinder already in the
sediment. After placing the dome on the cylinder, measurements were delayed for about
1 hr until the thermocouples approached thermal equilibrium with the surrounding water.
During each series of field measurements, water levels in the associated stream gauge and
piezometers were measured manually. When flux is low, signal response in the
21
-------
thermograph is feeble and less apparent as shown in Figure 20 as an example.
Nonetheless, the uprisings of both blue and cyan curves do suggest flow in the direction
of groundwater to surface water. The ascending of the green curve but not the red curve
reveals dispersion of heat up-gradient against flow direction, because at very low flows
dispersion of heat may occur in either direction. When this occurred, the active method
(addition or subtraction of known flow rate to the existing flow) could be employed to
improve the signal response for better data quality and also minimize the dispersion of
heat against flow direction, as shown in Figure 21.
Temper «
-------
The advective flux, in term of vertical Darcy velocity q in cm/day, measured at
each river cluster positions are reported in Table 1. Note that the analyses are based
strictly on peak temperature arrival and not moment analysis. Also reported in Table 1
are the vertical hydraulic head gradients (/', ft/ft) at each respective river cluster measured
manually on the same day when flux was measured. All measurements were made
between August 10 and September 20, 2005. Over all measurements, the range in the
flux was 0.3 - 2.63 cm/day. Table 1 shows values of flux measured with the passive
method were within 50% of each other, with 9 of 12 measurements occurring between
1.14 and 2.0 cm/day. This same variation in the measurements over spatial and temporal
scales also occurred using the active flow addition method; however the reproducibility
of the results at each location and time were greatly improved as indicated by the much
smaller standard deviations on these values.
The vertical hydraulic conductivity, K (cm/d), within the top 4 ft sediment layer is
calculated by dividing the value of flux by the corresponding vertical hydraulic head
gradient. The calculated values of K also are reported in Table 1 with the values
measured by the active method ranging from 1.55xlO"5 to 1.99><10"4 cm/sec, and the
values measured by the passive method ranging from 2.31 x 10"5 to 2.39x 10"4 cm/sec.
These values are within the range expected for the silty sedimentary material found at the
site.
Table 1. Advective flux (q), hydraulic conductivities (K), and vertical gradients (/') at each location. Data
from Dr. JafVert, Purdue University.
RC1
RC2N
RC2
RC2S
RC3
RC4
9/6/05
8/23/05
8/29/05
8/15/05
8/23/05
10/3/05
8/10/05
8/15/05
9/26/05
9/12/05
9/20/05
9/12/05
Passive Method
q K
(cm/day) at 0-4 ft (cm/sec)
1.90(0.85)a
1.89(0.30)
0.31 (0.038)
1.14(0.49)
1.66(0.38)
1.27(0.16)
2.63 (0.38)
2.00(0.13)
1.29(0.49)
1.46(0.42)
0.93 (0.40)
1.15(0.55)
l.SlxlO'4
1.44xlO'4
2.31xlO'5
8.52X10'5
1.45xlO'4
1.16xlO'4
2.39xlO'4
1.03xlO'4
5.89X10'5
2.12xlO'4
1.24xlO'4
1.33xlO'4
rf
3
11
3
3
6
3
9
3
3
3
3
3
Active Method
q K
(cm/day) at 0-4 ft (cm/sec)
1.49(0.07)a
0.78
0.36(0.05)
0.793 (0.02)
0.30 (0.26)
1.13(0.04)
1.50(0.20)
1.09(0.04)
1.03xlO'4
5.90xlO'5
2.70xlO'5
7.20 xlO'5
1.55xlO'5
5.21X10'5
1.99xlO'4
1.26xlO'4
n
3
1
3
3
2
3
3
3
i (ft/ft)
0-4 ft
0.168
0.153
0.155
0.155
0.133
0.128
0.128
0.225
0.253
0.080
0.088
0.100
4-8 ft
0.095
NDC
0.070
0.033
0.048
0.063
NDC
0.045
0.085
0.245
0.240
0.047
a Numbers in parenthesis are standard deviation of replicates.
b n = Number of measurements.
0 ND = Not determined, because the piezometric head in 8 ft piezometer did not reach at steady-
state.
23
-------
A final set of measurements at location RC2S were taken April 20-21, 2006. The
purpose of this data set was to compare data analysis methods between the method of
moment and peak temperature arrival. It was already known from previous
measurements that flow was likely from ground water to surface water and the flow
would be very small. RC2S location was chosen because the cylinder was previously
installed and left in the sediment. On April 20, the dome was re-attached to the cylinder,
and one set of data was taken to insure the unit was functional. Six data sets were
collected on the next day, three with passive method and three with active method where
5 ml/min were added.
Like the previous tests, flow sensor SN151 was also used; calibration curves shown
on Figure 8 & 11 were used instead to include moment analysis. Negative flow implies
groundwater discharge whereas positive flow indicates recharge. Figure 22 shows the
flux at RC2S estimated using both moment analysis and peak temperature arrival
methods. All data show groundwater discharge to the Grand Calumet River. The first 3
data sets in the earlier time scale were collected with the passive method, and the later 3
were collected with the active method. When the active method was employed, the
estimates based on the moment analysis yield smaller variation between the two channels
than the estimates based on the peak temperature arrival. However, there was insufficient
signal for meaningful moment analysis from the first and third set of data when the
passive method was used. Therefore, it is inconclusive to categorically say which data
analysis method is better.
Both passive and active methods produced similar results which gave confidence
that the flow bypassing was negligible at most; the loss of water from the aquifer was
consistently greater than zero, with 80% of the overall data showing upward groundwater
discharge between 0.37 to 1.21 cm/day. The overall moment analysis showed
groundwater discharge of 0.67+0.37 cm/day, while the peak arrival yielded 1.03+0.57
cm/day. Since one of the reasons for selecting this location for study was the
consideration of using a cap to isolate the contamination in the formation, the potential
for this low flow to dramatically reduce the life expectancy of a cap needs to be
investigated before implementation.
0.00
-0.50
-1.00
-1.50
-2.00
-2.50
4/21/06 4/21/06 4/21/06 4/21/06 4/21/06 4/21/06 4/21/06 4/21/06 4/21/06 4/21/06
7:40 8:09 8:38 9:07 9:36 10:04 10:33 11:02 11:31 12:00
Date & Time
Figure 22. Temporal flux at location RC2S estimated using both methods of analysis. Note: negative
flux value indicates groundwater discharge to surface water.
24
-------
3) Lake Hartwell, South Carolina
Monitored natural recovery (MNR) has being studied by the USEPA as a
potential remedy for the polychlorinated biphenyl (PCB) contaminated sediments at Lake
Hartwell. Advective flux through sediment could have potential negative impact on the
successful implementation of MNR. In May 2006, flux measurements were conducted at
Transect O and Transect N illustrated in Figure 23. Lake Hartwell provided an
opportunity to test the flux meter in deeper sediments in a large reservoir.
Measurements were made at two different locations. They are designated OB
(transect O location B) and NB (transect N location B). Locations OB and NB were
about 30 feet and 20 feet off the east shore, and the water depth was 14 feet and 10 feet,
respectively. The units tested at Lake Hartwell employed a two-piece dome with a
nominal Va" schedule 80 PVC pipe extended above water that functioned as a vent, but
also served as a mechanism to push the unit into sediment for installation. The cylinder
was polyethylene and the dome was constructed of stainless steel. Difficulties were
encountered in attempting to install the dome at depths of 10 feet+ in the turbid water.
The original thought was that the dome unit would be installed by a diver that would have
visual confirmation as to how the dome was contacting the sediment. However, a diver
was not available at the time. There was sufficient flex in the vent pipe as well as
movement in the boat leaving doubt as to whether the cylinder was properly sealed to the
bed sediment. Although a suction pull was felt when attempting retrieval, which could
serve as an indication that the dome was probably sealed to the bed sediment. There was
also flex in the dome itself because the stainless steel is thin and not very rigid. Where
this wasn't a problem in the shallow installations it proved to be problematic in the
deeper sediments especially when relying on pushing down the vent pipe for installation.
There was also potential for leaks at the seal between the cylinder and the dome, and the
connections between the dome and vent pipes.
Figure 23. Lake Hartwell Transect locations.
25
-------
The flux meter was first deployed at OB on May 3 in the early afternoon. Flow
sensor SN151 was used; calibration curves shown on Figure 8 & 11 were applied to the
data analysis. A few measurements were made to insure the unit was functional; the data
suggested groundwater recharge and the flux value ranged from 5.73 - 8.93 (median =
7.09) cm/day based on peak temperature arrival and 0.97 - 9.39 (median = 7.68) cm/day
based on the method of moment. Measurements were continued at the same location in
the morning of May 4. The temporal flux values averaged from the two channels are
presented in Figure 24 and Figure 25; results were based on the peak temperature arrival
and the method of moment, respectively. The flux meter was then re-deployed at NB in
the afternoon of May 4. During the deployments, there were boat activities near both
locations, especially quite frequent atNB, which were causing a lot of wave action and
disturbance to the water body during measurements. The unit as deployed only performs
adequately when the surface water is reasonably still. Some of the data appears to have
evidence of periodic flow in both directions (see Appendix F). It is believed that this is
likely due to wave action. It was possible that when the vent pipe remained open during
the measurements, hydraulic head could be fluctuated by sufficient lake seiche and/or
wave action, and alternated the flow direction making it impossible to interpret some of
the data.
Reasonable results were obtained at OB (see Figure 24 and 25), suggesting
groundwater recharge from surface water at the time of the measurements. Figure 24 and
25 were plotted on the same scale so the results can be visually compared. The temporal
variation of flux based on the peak temperature arrival appears to be smaller. The
variations between the two channels were also smaller based on the peak temperature
arrival. While the peak arrival analysis relies on one single peak temperature value, the
method of moment requires the entire longer period of data. Therefore, it is reasonable
that in an environment with noticeable wave action the method of moment could yield
higher uncertainty. Figure 24 shows the range in temporal groundwater recharge was
2.68-9.48 cm/day with an average of 5.53+2.44 cm/day. Figure 25 shows 11 out of 14
flux values were below 9 cm/day ranging 2.53-8.75 cm/day, with an average of
5.22+1.68 cm/day when discounting the three high values. Although sensible
interpretation of the data was impractical for NB due to poor data quality caused by the
frequent boat activities, the thermographs did confirm that the reservoir was loosing
water. Note that calibrations shown on Figure 8 & 11 were employed for the data
analysis. Positive flow, in this case, indicates groundwater recharge.
When the dome is first installed the flow system is disturbed. Measurements may
not correctly reflect the near steady state flow in the system. This has been observed at
each field location where measurements were made. Figure 24 & 25 both show higher
temporal variation in the early measurements than the later ones. The data illustrates the
concern expressed by (Shaw and Prepas 1989) regarding the impact of the installation of
the flow meter on the actual flow through the sediment. The data suggests that the
system required additional time to fully reach equilibrium. This demonstrates the
importance of having either semi-permanent deployment or the ability of returning to the
exact same spot as was done at Grand Calumet River in Hammond, Indiana. To reinstall
for deeper depths a diver must be used.
26
-------
The original design of the unit with the vent pipe open to the atmosphere is based
on the assumption that the changes in head during the measurement will be small. But
the data from Lake Hartwell suggested that much of the "noise" that was seen in the data
is likely head fluctuation as a result of wave action. In future measurement this can likely
be eliminated by using a closed dome that is periodically vented prior to measurement
sequence to remove gas accumulation, but is closed during the measurement. Some of
the standard addition or standard subtraction did not give the results anticipated. This
would suggest a leak in the system, or it could simply be owing to the inherent nature of
flow variation in a dynamic environment. There was no evidence that the flow meter was
not functional once it was returned to the laboratory.
1500
* * *
5/4/06 8:38 5/4/069:07 5/4/069:36 5/4/0610:04 5/4/0610:33 5/4/0611:02 5/4/0611:31 5/4/0612:00 5/4/0612:28
Date &Time
Figure 24. Temporal flux at location OB estimated based on peak temperature arrival. Note: positive
flux value indicates groundwater recharge from surface water.
1
5/4/068:38 5/4/069:07 5/4/069:36 5/4/0610:04 5/4/0610:33 5/4/0611:02 5/4/0611:31 5/4/0612:00 5/4/0612:28
Date & Time
Figure 25. Temporal flux at location OB estimated based on normalized first moments. Note: positive
flux value indicates groundwater recharge from surface water.
27
-------
Lessons Learned and Ideas for Improvement
1. From the field deployment in Nicaragua we learned that it would be desirable to
have a differential pressure transducer measuring the difference in pressure heads
in order to correct for the Bernoulli Effects. This will be most important when
installing the device where horizontal currents are significant.
2. It would be desirable to have a gauge pressure measurement that could monitor
the hydraulic head at the site. This will be important where the river/stream stage
is changing or there is wave action that could potentially change the flow.
3. We need to install a vent valve that can be opened or closed under program
control. This should minimize some of the problems that were observed where
the surface water was turbulent. This will be important whenever the head
changes significantly during the measurement period.
4. When installing the dome at deeper locations it would be preferable to utilize a
diver. If this is not practical then an underwater video camera should be used to
monitor the installation process. The vent pipe needs to be made out of a heavy
gauge metal to reduce flex during installation. This is particularly important if the
installation is going to be from the surface rather than by a diver. Currently there
is also too much flex in the dome. It is recommended that the dome be
constructed out of a more robust material.
5. The location of the flow sensor also needs to be considered as well as its
orientation. If the dome is installed from the surface, there is concern that
suspended sediment might be pushed up and accumulated in the flow sensor
creating errors in the measurements or creating a potential pressure drop across
the device leading to measurement errors. It should be considered to first install
the dome and flush the dome before installing the flow sensor.
6. Multi-elevation piezometer data will be needed if calculating the hydraulic
conductivity of the bed sediment is advantageous. For evaluating the impacts of
capping this will be beneficial. It would also be desirable to measure the vector of
the potential gradient. The flow meter only measures the flow normal to the plane
of the collection cylinder.
7. There is electronic noise that is being picked up due to the very low level signals
being transmitted. It is believed that these signals should be converted to a
frequency output at the signal source so that attenuation and cross talk is
minimized.
8. Temporal data either returning to the same location or leaving the unit at the same
location will permit learning about more long term changes. The data at the
Hammond Indiana site illustrate a near continuous record. It is believed
continuous monitoring will be required to identify and avoid unsound readings
due to perturbations caused by the initial installation. Short term measurements
should only be considered as qualitative rather than quantitative.
28
-------
9. For long term monitoring, it would be desirable to transmit the signal from a buoy
located at the water surface back to land where it could be re-transmitted to an
office environment for continuous monitoring. This would require converting the
system to one that could be operated from solar cell recharged batteries and a
wireless network installation.
10. At the present time the unit does not allow for the collection of water samples.
When studying water quality issues, it would be desirable to be able to collect
representable samples discharging form the sediment. It would have problems
properly sampling where the flow direction reverses frequently.
Conclusions
Significant advancements have been achieved in the development and
demonstration of a bidirectional advective flux meter. We have successfully developed
the device based on a heat-pulse technique and derived empirical relations between flow
rate and heat-pulse travel time. Operational control and data analysis software were
developed in Visual Basic 6.0. Both peak temperature arrival method and the method of
first temporal moment were made available to provide multiple data analysis options.
Each analysis approach has advantages depending on the environmental conditions when
the measurements were made. Laboratory calibrations of both methods yield quality
results with R2 of 0.99 or better. The flux meter has undergone field tests at three
different sites. The first test site was a shallow turbulent stream at Santo Domingo,
Nicaragua; the large swings in hydrograph caused by the frequent rainfall events during
the field deployment prevented us from obtaining reproducible data. The second test site
was at Grand Calumet River, Hammond, IN, which is a slow moving river with fine
sediments. And the final test site was at a large reservoir with deeper sediments at Lake
Hartwell, SC. The last two field deployments were successful in that the flux meter was
operational, reasonable characterization of flux were obtained, and bidirectional flow
measurement capability was demonstrated. The field demonstration has provided insights
on how improvements can be made. Nevertheless, under appropriate conditions the
units, as currently configured, should provide continued long term data and analysis
capabilities.
Acknowledgement
This work couldn't have been completed without the goodwill and efforts of Dr.
Goran Bengtsson, University of Lund Sweden, Dr. Chad Jafvert, Purdue University and
Dr. Marc Mills, U.S. EPA. They have been instrumental in providing the sites where the
unit was tested and offering continued encouragement to the development and testing of
the device. The author also likes to express his gratitude to the University of Cincinnati
contract support for providing invaluable contribution and innovation to this project.
29
-------
References
Annable, M. D., K. Hatfield, et al. (2005). "Field-Scale Evaluation of the Passive Flux
Meter for Simultaneous Measurement of Groundwater and Contaminant Flux."
Environ. Sci. and Technol. 39: 7194-7201.
Basinger, J. M., G. J. Kluitenberg, et al. (2003). "Laboratory evaluation of the dual-probe
heat pulse method for measuring soil water content." Vadose Zone J 2: 389-399.
Belanger, T. V. and M. T. Montgomery (1992). "Seepage mete errors." Limnology and
Oceanography 37(8): 1787-1795.
Bilskie, J. R., R. Horton, et al. (1998). "Test of a dual-probe heat-pulse method for
determining thermal properties of porous materials." Soil Sci. 163: 346-355.
Bockelmann, A., T. Ptak, et al. (2001). "An analytical quantification of mass fluxes and
natural attenuation rate constants at a former gasworks site." Contaminant
Hydrology 53: 429-453.
Campbell, G. S., C. Calissendorff, et al. (1991). "Probe for measuring soil specific heat
using heat-pulse method." Soil Sci Am J 55: 291-293.
Choi, H.-m., K.-B. Lee, et al. (1991). "Measurement of gas flowrate by using a heat
pulse." Flow Measurement and Instrumentation 2(3): 188.
de Jonge, H. and G. Rothenberg (2005). "New Device and Method for flux-Proportional
Sampling of Mobile Soluites in Soil and Groundwater." Environ. Sci. and
Technol. 39: 274-282.
Fellows, C. r. and P. L. Brezonik (1980). "Seepage flow into Florida lakes." Water
Resources Bulletin 16(4): 635-641.
Harris, P. J. (1995). "Water quality impacts from on site waste disposal systems to coastal
areas through groundwater discharge." Environmental Geology 26(4): 262-268.
Heitman, J. L., J. M. Basinger, et al. (2003). "Field evaluation of the dual-probe heat-
pulse method for measuring soil water content." Vadose Zone Journal 2(4): 552-
560.
Huettel, M., W. Ziebis, et al. (1998). "Advective Transport Affecting Metal and Nutrient
Distributions and Interfacial Fluxes in Permeable Sediments." Geochimica et
Cosmochimica Acta 62(4): 613-631.
Jafvert, C. T., D. Lane, et al. (2006). "Partitioning of mono- and polycyclic aromatic
hydrocarbon in a river sediment adjacient to a former manufactured gas plant
site." Chemosphere 62: 315-321.
Jahnke, R. A., C. R. Alexander, et al. (2003). "Advective pore water input of nutrients to
the Satilla River Estuary, Georgia, USA." Estuarine, Coastal and Shelf Science
56(3-4): 641-653.
Jahnke, R. A., J. R. Nelson, et al. (2000). "Benthic flux of biogenic elements on the
Southeastern US continental shelf: influence of pore water advective transport and
benthic microalgae." Continental Shelf Research 20(1): 109-127.
Kalbus, E., F. Reinstorf, et al. (2006). "Measuring methods for groundwater,surface water
and their interactions: a review." Hydrol. Earth Syst. Sci. Discuss. 3: 1809-1850.
Krupa, S. L., T. V. Belanger, et al. (1998). "Krupaseep - the next generation seepage
meter." Journal of Coastal Research SI 26: 210-213.
Lee, D. R. (1977). "A device for measuring seepage flux in lakes and estuaries."
Limnology and Oceanography 22(1): 140-147.
30
-------
Lee, D. R. and J. A. Cherry (1978). "A field exercise on groundwater flow using seepage
meters and mini-piezometers." Journal of Geological Education 27(6-20).
Li, L. and D. A. Barry (2000). "Wave-induced beach groundwater flow." Advances in
Water Resources 23: 325-337.
Li, L., D. A. Barry, et al. (1999). "Submarine groundwater discharge and associated
chemical input to a coastal sea." Water Resources Research 35(11): 3253-3259.
Libelo, E. L. and W. G. Maclntyre (1994). "Effects of surface-water movement on
seepage meter measurements of flow through the sediment-water interface."
Applied Hydrogeology 2: 49-54.
Liu, C. H., J. A. Jay, et al. (2001). "Evaluation of environmental effects on metal
transport from capped contaminated sediment under conditions of submarine
groundwater discharge." Environmental Science & Technology 35(22): 4549-
4555.
Menheer, M. A. (2004). Development of a benthic-flux chamber for measurement of
ground-water seepage and water sampling for mercury analysis at the sediment-
water interface. U.S. Geological Survey Scientific Investigation Report 2004-
5298. U.S. Geological Survey.
Mortensen, A. P., J. W. Hopmans, et al. (2006). "Multi-functional heat pulse probe
measurements of coupled vadose zone flow and transport." Advances in Water
Resources 29(2): 250.
Murdoch, L. C. and S. E. Kelly (2003). "Factors affecting the performance of
conventional seepage meters." Water Environ. Res. 39(6): 1163.
Paillet, F., R. Crowder, et al. (1996). "High-Resolution Flowmeter Logging Applications
with the Heat-Pulse Flowmeter." Journal of Environmental and Engineering
Geophysics 1(1): 1-11.
Paulsen, R. J., C. F. Smith, et al. (2001). "Development and evaluation of an ultrasonic
ground water seepage meter." Ground Water 39(6): 904-911.
Precht, E. and M. Huettel (2004). "Rapid wave-driven advective pore water exchange in a
permeable coastal sediment." Journal of Sea Research 51(2): 93-107.
Ren, T., G. J. Kluitenberg, et al. (2000). "Determining soil water flux and pore water
velocity by a heat pulse technique." Soil Sci Am J 64: 552-560.
Rosenberry, D. O. and R. H. Morin (2004). "Use of an electromagnetic seepage meter to
investigate temporal variability in lake seepage." Ground Water 42(1): 68-77.
Schincariol, R. A. and J. D. McNeil (2002). "Errors with small volume elastic seepage
meter bags." Ground Water 40(6): 649-651.
Sebestyen, S. D. and R. L. Schneider (2001). "Dynamic temporal patterns pf nearshore
seepage flux in a headwater Adirondack lake." Journal of Hydrology 247: 137-
150.
Shaw, R. D. and E. E. Prepas (1989). "Anomalous, short-term influx of water into seeage
meters." Limnology and Oceanography 34(7): 1343-1351.
Shaw, R. D. and E. E. Prepas (1990). "Groundwater-lake interactions, I: Accuracy of
seepage meter setimates of lake seepage." Journal of Hydrology 119: 105-120.
Shaw, T. J., W. S. Moore, et al. (1998). "The flux of barium to the coastal waters of the
southeastern USA: The importance of submarine groundwater discharge."
Geochimica Et Cosmochimica Acta 62(18): 3047-3054.
-------
Shinn, E. A., C. D. Reich, et al. (2002). "Seepage meters and Bernoulli's revenge."
Estuaries 25(1): 126-132.
Sholkovitz, E., C. Herbold, et al. (2003). "An automated dye-dilution based seepage
meter for the time-series measurement of submarine groundwater discharge."
Limnology and Oceanography Methods 1: 16-28.
Shook, G. M. and J. H. Forsmann (2005). Tracer interpretation using temporal moments
on a spreadsheet, Idaho National Laboratory. INL/EXT-05-00400.
Simmons, G. M. (1992). "Importance of Submarine Groundwater Discharge (Sgwd) and
Seawater Cycling to Material Flux Across Sediment Water Interfaces in Marine
Environments." Marine Ecology-Progress Series 84(2): 173-184.
Slomp, C. P. and P. Van Cappellen (2004). "Nutrient inputs to the coastal ocean through
submarine groundwater discharge: controls and potential impact." Journal of
Hydrology 295(1-4): 64-86.
Sonnenschmidt, D. and K. H. Vanselow (1996). "Measurement of water flow velocity by
analysis of short temperature steps." Meas. Sci. Technol. 7: 1536-1539.
Taniguchi, M., W. C. Burnett, et al. (2003). "Spatial and temporal distributions of
submarine groundwater discharge rates obtained from various types of seepage
meters at a site in the Northeastern Gulf of Mexico." 66(1-2): 35.
Taniguchi, M. and Y. Fukuo (1993). "Continuous Measurements of Groundwater
Seepage Using an Automatic Seepage Meter." Ground Water 31(4): 675-679.
Taniguchi, M. and Y. Fukuo (1996). "An effect of seiche on groundwater seepage rate
into Lake Biwa, Japan." Water Resour. Res. 32: 333-338.
U.S. EPA (2000). Proceedings of the Ground-Water/Surface-Water Interactions
Workshop. EPA 542/R-00/007. Washington, DC, Office of Solid Waste and
Emergency Response.
U.S. EPA (2006). "Evaluating ground-water/surface-water transition zones in ecological
risk assessments." ECO Update/Ground Water Forum Issue Paper in final Office
review and approval
Uchiyama, Y., K. Nadaoka, et al. (2000). "Submarine groundwater discharge into the sea
and associated nutrient transport in a sandy beach." Water Resources Research
36(6): 1467-1479.
Winter, T. C., J. W. LaBaugh, et al. (1988). "The design and use of a hydraulic
potentiomanometer for direct measurement of differences in hydraulic head
between groundwater and surface water." Limnology and Oceanography 33(5):
1209-1214.
Wu, X. (2006). An investigation of partitioning tracers for characterizing geothermal
reservoirs and predicting enthalpy production, The University of Texas at Austin:
277.
Yang, H.-S., D.-W. Hwang, et al. (2002). "Factors controlling excess radium in the
Nakdong River estuary, Korea: submarine groundwater discharge versus
desorption from riverine particles." Marine Chemistry 78(1): 1-8.
32
-------
Appendix A Calibration Curve for SN152
Calibration for sensor # 152
ChO
Ch 1
Fit Ch 1
Fit Ch 0
Ch3
On 4
-Fit Ch 3
- Fit Ch 4
200
300 400 500
Measured Time (sec)
600
700
eon
33
-------
Appendix B Calibration Curve for SN151
Heating cycle = 3 seconds
SN151 Flux Meter
• Channel 0 y = 3688.2X-1258r R2 = 0.9842
A Channel 1 y = 4037.6x-14251 R2 = 0.9794
A Channel 3 y = 3260.3x'1368 R2 = 0.9906
o Channel 4 V = 2240.5x~1 1442 R2 = 0.9854
25 75 125 175 225 275 325 375
Heat Pulse Arrival Time (sec)
425 475
34
-------
Appendix C Computer Code
Option Explicit
' Scans a range of A/D Input Channels, stores the data in an array
' and output the data to Ascii files.
'Averages signals collected to minimize noise by scannng the data _
before output to file
'send digital output to switch heater ON/OFF
'channel 0-1: flow temperature
'channel 2: heater temperature
'channel 3-4: flow temperature
'channel 5: reference junction temperature
'channel 6: Unique resistor used to determine serial number.
'channel 7: Internal Reserved for future use most likely a pressure transducer
' last revised 02/1/06 by Carl Enfield
Const BoardNum% = 1 ' Board number
Const PortNum% = AUXPORT ' use digital auxilary port
Const Direction% = DIGITALOUT ' program digital auxilary port for output
Const FirstPoint& = 0 'set first element in buffer to transfer to array
Dim ADData%() ' dimension an array to hold the input values
Dim addatal(S) ' array to hold initial thermocouple offset data
Dim addata2() 'dimension an array to transer ADData% to double precision numbers
Dim addata3(8)
Public baseline 'time used to collect temperature data prior to injecting heat
Public heater 'time in seconds heater is turned on
Public tail ' time in seconds experimental data is collected
Public pause ' time in seconds before next cycle of data collection allos heat to
dissipate
Dim counter%, rounds, RateCount
Dim DataValue% ' digital value to write to Auxport
Dim fname As String * 50
Dim fname2 As String * 50
Dim site As String * 64
Public LowChan% 'First cahnnel to read
Dim MemHandle& ' define a variable to contain the handle for
' memory allocated by Windows through cbWinBufAlloc%()
Public HighChan% ' Last channel to be read
Dim i, k, j, N ' counter
Dim Options ' used as variable in the scan function call
Dim SampleTime, startpause
35
-------
Dim PCCDAS1616AO, mytime
Dim finish, timepast, t, timezero
Dim sum(8)
Dim ULStat As Long 'universal library styatistic return
Dim voltrange
Dim SumAdata As Double
Dim SRate As Long 'Scan rate universal library function
Dim DataQ
Dim A_Data As Double
Dim DomeDia
Dim SerialNo
Dim textline
Dim ResistorQ, clOQ, c20(), c30(), cvOQ, cll(), c21(), _
c31(), cvl(), cl3(), c23(), c33(), cv3(), cl4(), c24(), _
c34(), cv4(), SerNoQ, TslopeQ, TintQ
Private Sub CmdClose_Click()
Unload Me
End Sub
Private Sub cmdStart_Click()
i*******
' Turn off heater
SRate = 10000
DataValue% = 1
'config. digital output port
ULStat = cbDConfigPort(BoardNum%, PortNum%, Direction%)
If ULStat <>0 Then Stop
'digital out
ULStat = cbDOut(BoardNum%, PortNum%, DataValue%)
If ULStat <>0 Then Stop
' Determine serial number of flow sensor by measuring voltage on channel 6.
' Look up flow meter in data base and make sure it exists then display serial
' number on display
ReDim ADData%(100)
ReDimData(lOO)
MemHandle& = cbWinBufAlloc(lOO)
If MemHandle& = 0 Then Stop
Date = Now
Time = Now
36
-------
baseline = CDbl(txtBaseline.Text) ' time to collect baseline data (sec)
heater = CDbl(txtHeater.Text) ' time for heater to be turned on (sec)
DomeDia = CDbl(txtDDia.Text) ' Diameter of the Dome
voltrange = BIP 1 0 VOLTS ' voltage range +- 1 0 VDC
site = CStr(txtSite.Text) ' site description
ULStat = cbAInScan(BoardNum%, 6, 6, 100, SRate, voltrange, MemHandle&,
Options)
If ULStat = 91 Then
ULStat = cbErrHandling(DONTPRINT, DONTSTOP)
Elself ULStat <> 0 Then
Stop
End If
' transfer values from Windows buffer to data array used by VB A
ULStat = cbWinBufToArray(MemHandle&, ADData%(l), FirstPoint&, 100)
ULStat = cbWinBufFree(MemHandle&) ' Free up memory for use by
If ULStat <> 0 Then Stop ' other programs
SumAdata = 0
Forj = 1 To 100
ULStat = cbToEngUnits(BoardNum%, voltrange, ADData%(j), Data(j))
SumAdata = SumAdata + Data(j)
Nextj
A_Data = SumAdata / (100# * 0.0001) 'assumes a current source of 100 uA is used to
evaluate resistor
'write baseline, heat-pulse duration, Resistance, and Site description
fname = "c:\temp\" + Format$(Date, "mmm dd ") + Format$(Time, "hh mm ss") +
".dat"
Open fname For Output As #1
Write #1, Format(Now), Format(baseline, "#0.0"), _
Format(heater, "#0.0"), Format(A_Data, "#0.0000"), _
Format(DomeDia, "#0.0000"), Format(site)
Close
' Look up calibration data and display sensor serial number
ChDir ("C:\Program Files\Flux_Meter_CGE")
Open "Calibration_Data.csv" For Input As #4
N = 0
Do While Not EOF(4)
Line Input #4, textline
N = N + 1 'n is the number of records or values
Loop
Close #4
37
-------
'MsgBox (n)
'Allocate space to hold arrays place calibration data into arrays
ReDim Resistor(N), clO(N), c20(N), c30(N), cvO(N), cll(N), c21(N), _
c31(N), cvl(N), c!3(N), c23(N), c33(N), cv3(N), c!4(N), c24(N), _
c34(N), cv4(N), SerNo(N), Tslope(N), Tint(N)
ReDim Resistor(N), conO(N), expO(N), conl(N), expl(N), con3(N), exp3(N), con4(N),
exp4(N), SerNo(N)
Open "Calibration_Data.csv" For Input As #5
N = 0
Do While Not EOF(5)
Input #5, Resistor(N), clO(N), c20(N), c30(N), cvO(N), cll(N), c21(N), _
c31(N), cvl(N), c!3(N), c23(N), c33(N), cv3(N), c!4(N), c24(N), _
c34(N), cv4(N), SerNo(N), Tslope(N), Tint(N)
N = N+ 1
Loop
Close #5
' Find serial Number assumption value can be detected +-2% +-3 ohms
Forj=2ToN- 1
If A_Data > (Resistor(j) * 0.98 - 3) And _
A_Data < (Resistor(j) * 1.02 + 3) Then GoTo 100
Nextj
If j = N Then SerNo(j - 1) = "default" ' Serial number not detected
j=j-l
100 IblSerNo. Caption = SerNoG)
MemHandle& = cbWinBufAlloc(lOO) ' begin scan for cahnnel 0
For i = 1 To 5
ULStat = cbAInScan(BoardNum%, i - 1, i - 1, 100, SRate, voltrange, MemHandle&,
Options)
If ULStat = 91 Then
ULStat = cbErrHandling(DONTPRINT, DONTSTOP)
Elself ULStat <> 0 Then
Stop
End If
' transfer values from Windows buffer to data array used by VB A
ULStat = cbWinBufToArray(MemHandle&, ADData%(l), FirstPoint&, 100)
SumAdata = 0
Forj = 1 To 100
ULStat = cbToEngUnits(BoardNum%, voltrange, ADData%(j), Data(j))
38
-------
SumAdata = SumAdata + Data(j)
Nextj
addatal(i) = SumAdata / (100#)
Next i
ULStat = cbWinBufFree(MemHandle&) ' Free up memory for use by
If ULStat <> 0 Then Stop ' other programs
' Initialize settings required to scan thermocouples
' Chanels 0 through 5 are scanned. All range settings are set the same. Each channel
' is measured 100 times using the cbAInScan function.
MemHandle& = cbWinBufAlloc(lOO)
If MemHandle& = 0 Then Stop
ReDim ADData%(100)
tail = CDbl(txtTail.Text) ' time to collect data avter heater is turned off (sec)
pause = CDbl(txtPause.Text) ' wait time before next sampling round (sec)
' S_rate = CDbl(txtRate.Text) ' sampling rate (samples/sec)
rounds = CDbl(txtRounds.Text) ' Number of data sets to collect
counter% = 0 'sampling round counter
SampleTime = baseline + heater + tail 'total sampling time
voltrange = BIP1 VOLTS ' voltage range +- 1 VDC
Do While PCCD AS 1 6 1 6 AO <> - 1 'main loop
Date = Now
Time = Now
counter% = counter% + 1
Lab el 12. Caption = counter%
timezero = Timer
t = 0
k = 0
1 1 Do While t < SampleTime - 1
'calculate accmulated time
If Timer < timezero Then 'if cross midnight
t = 86400 - timezero + Timer
Else
t = Timer - timezero
End If
39
-------
'control when to activate analog output
If t <= baseline Then DataValue% = 1: txtBaseline.BackColor = &HFFOO&
If t > baseline Then txtBaseline.BackColor = &HFFFFFF
If (t > baseline And t <= (baseline + heater)) Then DataValue% = 0:
txtBaseline.BackColor = &HFFFFFF: txtHeater.BackColor = &HFFOO&
If t > (baseline + heater) Then DataValue% = 1: txtHeater.BackColor = &HFFFFFF:
txtTail.BackColor = &HFFOO&
'digital out
ULStat = cbDOut(BoardNum%, PortNum%, DataValue%)
If ULStat <> 0 Then Stop
MemHandle& = cbWinBufAlloc(lOO) ' begin scan for cahnnel 0
For i = 1 To 6
ULStat = cbAInScan(BoardNum%, i - 1, i - 1, 100, SRate, voltrange, MemHandle&,
Options)
If ULStat = 91 Then
ULStat = cbErrHandling(DONTPRINT, DONTSTOP)
Elself ULStat <> 0 Then
Stop
End If
' transfer values from Windows buffer to data array used by VB A
ULStat = cbWinBufToArray(MemHandle&, ADData%(l), FirstPoint&, 100)
SumAdata = 0
Forj = 1 To 100
ULStat = cbToEngUnits(BoardNum%, voltrange, ADData%(j), Data(j))
SumAdata = SumAdata + Data(j)
Nextj
addata3(i) = SumAdata / (100#)
Next i
ULStat = cbWinBufFree(MemHandle&) ' Free up memory for use by
If ULStat <> 0 Then Stop ' other programs
i
Open fname For Append As #1
'write data in voltage
Write #1, Format(Now), Format(t, "#0.00000"), _
Format(addata3(l), "#0.00000"), Format(addata3(2), "#0.00000"), _
Format(addata3(3), "#0.00000"), Format(addata3(4), "#0.00000"), _
Format(addata3(5), "#0.00000"), Format(addata3(6), "#0.00000"), _
Format(addata3(7), "#0.00000"), Format(addata3(8), "#0.00000")
Close
'display data in voltage
For i = 1 To 5
IblADData(i).Caption = Format((addata3(i) - addatal(i)) * 10, "#0.000")
Next i
lblADData(6).Caption = Format(addata3(6) * 100, "#0.00")
40
-------
lblADData(9). Caption = Now 'display time
Refresh
DoEvents
If PCCDAS1616AO = -1 Then GoTo 44
Loop '11
48 Close #1
txtTail.BackColor = &HFFFFFF
startpause = Timer ' Set start time.
t = 0
txtPause.BackColor = &HFFOO&
14 Do While t< pause
If Timer < startpause Then 'if cross midnight
t = 86400 - startpause + Timer
Else
t = Timer - startpause
End If
DoEvents ' Yield to other processes
If PCCDAS1616AO = -1 Then GoTo 44
Loop '14
txtPause.BackColor = &HFFFFFF
' lblADData(8).BackColor = &HFFFFFF
If counter% = rounds Then PCCDAS1616AO = -1
If (counter% < rounds) Then
fname = "c:\temp\" + Format$(Date, "mmm dd ") + Format$(Time, "hh mm ss") +
".dat"
Open fname For Output As #1
Write #1, Format(Now), Format(baseline, "#0.0"), _
Format(heater, "#0.0"), Format(A_Data, "#0.0000"), _
Format(DomeDia, "#0.0000"), Format(site)
Close
End If
Loop 'main
'PCCD AS 1616 AO = 0 'continue
44 txtBaseline.BackColor = &HFFFFFF
txtHeater.BackColor = &HFFFFFF
41
-------
txtTail.BackColor = &HFFFFFF
txtPause.BackColor = &HFFFFFF
cmdStart.Enabled = -1
PCCDAS1616AO = 0
GoTo 45
45 End Sub
Private Sub CmdStop_Click()
PCCDAS1616AO = -1
End Sub
Private Sub cmdStopConvert_Click()
ULStat = cbWinBufFree(MemHandle&) ' Free up memory for use by
' other programs
1 If ULStat <>0 Then Stop
End
End Sub
Private Sub Form_Load()
' declare revision level of Universal Library
ULStat = cbDeclareRevision(CURRENTREVNUM)
' Initiate error handling
' activating error handling will trap errors like
' bad channel numbers and non-configured conditions.
' Parameters:
' PRINTALL :all warnings and errors encountered will be printed
' DONTSTOP :if an error is encountered, the program will not stop,
' errors must be handled locally
ULStat = cbErrHandling(PRINTALL, DONTSTOP)
If ULStat <>0 Then Stop
1 If cbErrHandling% is set for STOP ALL or STOPFATAL during the program
' design stage, Visual Basic will be unloaded when an error is encountered.
' We suggest trapping errors locally until the program is ready for compiling
' to avoid losing unsaved data during program design. This can be done by
' setting cbErrHandling options as above and checking the value of ULStat%
42
-------
after a call to the library. If it is not equal to 0, an error has occurred.
End Sub
Private Sub Initialize_Click()
DataValue% = 0
'config. digital output port
ULStat = cbDConfigPort(BoardNum%, PortNum%, Direction%)
If ULStat <> 0 Then Stop
'digital out
ULStat = cbDOut(BoardNum%, PortNum%, DataValue%)
If ULStat <> 0 Then Stop
End Sub
43
-------
'last revised 02/1/06 by Carl Enfield
Option Explicit
Private Sub CmdClose_Click()
Unload Me
End Sub
Private Sub Dirl_Change()
Filel.Path = Dirl
End Sub
Private Sub Form_Load()
Dirl.Path= "c:\temp"
Filel.Path = Dirl
End Sub
Private Sub Commandl_Click() ' Listl
Screen.MousePointer =11
Dim textline
Dim timeOQ As String
Dim timel As String 'used to store time file created needed to determine age of file and
be compatable with older data
Dim ch() 'dynamic array to hold analog data stored on .dat file
Dim t() 'dynamic array to hold elapsed time data
Dim ct As Integer ' , pt_odd, pt_half As Integer
Dim tmax(6) As Double
Dim tempmax(6)
Dim bline
Dim heater
Dim N, j, i, 1, m, g, k, z, ii, Nl As Integer
Dim CalResistor
Dim ResistorQ
Dim DomeDia
Dim site As String * 64
Dim QO, Ql, Q3, Q4, QOmin, Qlmin, Q3min, Q4min, QOmax, Qlmax, Q3max,
Q4max
Dim VO, VI, V3, V4
Dim dat_type As Boolean
ChDir Dirl
' Initialize all of the calculated outputs other than peak arrival time to "N.A."
' to make sure only recomputed values are displayed if this isn't done
' it is possibe to display a value from a previous data set
IblMoml. Caption = Format("N.A.")
44
-------
lblMom2.Caption = Format("N.A.")
lblMom4. Caption = Format("N.A.")
IblMomS. Caption = Format("N.A.")
QO = "N.A."
QOmin="N.A."
QOmax="N.A."
IblQO.Caption = Format(QO)
IblQOmin. Caption = Format(QOmin)
IblQOmax. Caption = Format(QOmax)
Q1 = "N.A."
Qlmin="N.A."
Qlmax="N.A."
IblQl. Caption = Format(Ql)
IblQlmin. Caption = Format(Qlmin)
IblQl max. Caption = Format(Qlmax)
Q3 = "N.A."
Q3min="N.A."
Q3max="N.A."
lblQ3. Caption = Format(Q3)
lblQ3min. Caption = Format(Q3min)
lblQ3max. Caption = Format(Q3max)
Q4 = "N.A."
Q4min="N.A."
Q4max="N.A."
lblQ4.Caption = Format(Q4)
lblQ4min. Caption = Format(Q4min)
lblQ4max. Caption = Format(Q4max)
VO = "N.A."
VI = "N.A."
V3 = "N.A."
V4 = "N.A."
IblVO.Caption = Format(VO)
IblVl. Caption = Format(Vl)
lblV3. Caption = Format(V3)
lblV4.Caption = Format(V4)
' Import calibration data
ChDir ("C:\Program Files\Flux_Meter_CGE\")
Open "Calibration_Data.csv" For Input As #6
i******
' determine the number of calibration curves in the data table
45
-------
1 = 0
Do While Not EOF(6)
Line Input #6, textline
1 = 1+1 '1 is the number of calibration curves
Loop
Close #6
'Allocate space to hold arrays place calibration data into arrays
ReDim Resistor(l), clO(l), c20(l), c30(l), cvO(l), cll(l), c21(l), _
c31(l), cvl(l), c!3(l), c23(l), c33(l), cv3(l), c!4(l), c24(l), _
c34(l), cv4(l), SerNo(l), Tslope(l), Tint(l)
Open "Calibration_Data.csv" For Input As #7
' Import calibration data
1 = 1
Do While Not EOF(7)
Input #7, Resistor(l), clO(l), c20(l), c30(l), cvO(l), cll(l), c21(l),
c31(l), cvl(l), c!3(l), c23(l), c33(l), cv3(l), c!4(l), c24(l), _
c34(l), cv4(l), SerNo(l), Tslope(l), Tint(l)
1 = 1 + 1
Loop
1 MsgBox (1)
Close #7
ChDir Dirl
'Determine the number of records in the data file
Open Filel For Input As #1
Nl = 0
N = 0
Do While Not EOF(l)
Line Input #1, textline
N = N + 1 'n is the number of records or values
Nl = Nl + 1 'Nl is the number of records or values do not reset this counter
Loop
Close #1
If(Nl -20<0)Then
MsgBox "Not enough data points. Please check file."
MsgBox (tO))
46
-------
MsgBox (bline)
GoTo 44
End If
'Allocate space to hold arrays
ReDim timeO(N + 1) As String '+1?
ReDim t(N)
ReDim ch(j,N)
' find data collection date older data did not have a header file with information
' on time of the base line or time heater was on
Open Filel For Input As #3
Input #3, timel, bline, heater, CalResistor, DomeDia, site
Close #3
before 4/26/04 there was no header line on the data file. From4/26/04 to 8/4/05
there was a header with three values the time the measurement was initiated,
the time the base line data was collected, and the time the heater was turned on.
After 8/4/05 three additional parameters were added the estimate of the resistor
included in the to select the calibration curve, the diameter of the dome, and
a field to identify the purpose of the data collection.
If (Year(timel) < 2004) Then GoTo 3
If (Year(timel) = 2004 And Month(timel) < 4) Then GoTo 3
If (Year(timel) = 2004 And Month(timel) = 4 And Day(timel) <= 25) Then GoTo
3 'pre header file
If (Year(timel) < 2005) Then GoTo 4
If Year(timel) = 2005 And Month(timel) < 8 Then GoTo 4
If (Year(timel) = 2005 And Month(timel) = 8 And Day(timel) <= 4) Then GoTo 4
If (Year(timel) = 2005 And Month(timel) = 8 And Day(timel) <= 4) Then GoTo 4
If (IsDate(CalResistor) = True) Then GoTo 4
GoTo 5
Open Filel For Input As #2 ' place data in arrays
j = 0
N = 0
bline = 0
heater = 0
Do While Not EOF(2)
j=j + l
N = N+ 1
Input #2, timeOG), tQ), ch(l, j), ch(2, j), ch(3, j), ch(4, j), ch(5, j), ch(6, j), ch(7, j), ch(8,
j)
47
-------
Loop
Close #2
IblSiteD. Caption = "No Data" 'No Data available
IblSerNo. Caption = "Not Assigned" 'Before Serial numbers
GoTo 14
4 Open Filel For Input As #4
' MsgBox ("branch 4")
j = 0
N = 0
Do While Not EOF(4)
If j = 0 Then
Input #4, timeO(j), bline, heater
End If
j=j + l
N = N+ 1
Input #4, timeOG), tQ), ch(l, j), ch(2, j), ch(3, j), ch(4, j), ch(5, j), ch(6, j), ch(7, j), ch(8,
j)
Loop
Close #4
IblSiteD.Caption = "No Data" 'No Data available
IblSerNo. Caption = "Default" 'Serial Number data not stored in file default used
GoTo 14
5 Open Filel For Input As #5
j = 0
N = 0
Do While Not EOF(5)
If j = 0 Then
Input #5, timeO(j), bline, heater, CalResistor, DomeDia, site
'MsgBox (bline)
'MsgBox (heater)
'MsgBox (CalResistor)
End If
j=j + l
N = N+ 1
Input #5, timeOG), tQ), ch(l, j), ch(2, j), ch(3, j), ch(4, j), ch(5, j), ch(6, j), ch(7, j), ch(8,
j)
Loop
Close #5
IblSiteD.Caption = site
' Find serial Number assumption value can be detected +-2% +-3 ohms
For i = 2 To 1 - 1
If CalResistor - (Resistor(i) * 0.98 - 3) > 0 And _
48
-------
CalResistor - (Resistor(i) * 1.02 + 3) < 0 Then GoTo 100
Nexti
If i = 1 Then SerNo(i - 1) = "default" ' Serial number not detected
i = i- 1
100 IblSerNo. Caption = SerNo(i)
ii = i' ii is used as the record number of the for the calibration data
'MsgBox (ii)
' Determine if sufficient data for possible analysis
14 If(t(j)-bline<0)Then
MsgBox "Not enough data points. Please check file."
MsgBox (tO))
' MsgBox (bline)
GoTo 44
End If
' Analyze base line data and determine slope and intercept analysis based on baseline
' data and last 30 observations. These values are averaged to calculate the slope and
' intercept assuming a streight line is appropriate. There should probably be a data
' string length of at least 3 times the peak arival point.
Dim sumx(6)
Dim sumy(6)
Dim sumxl(6)
Dim sumyl(6)
Dim avgx(6)
Dim avgy(6)
Dim Slope(6)
Dim slnt(6)
Dim avgxl(6)
Dim avgyl(6)
'calculate sums used to calculate baseline averages, slopes, and intercepts
For j = 1 To 6
sumx(j) = 0
sumy(j) = 0
sumxl(j) = 0
sumyl(j) = 0
Nextj
MsgBox (bline)
ct = 1 'counter
Do 'While t(ct) < bline
49
-------
' MsgBox ((t(ct) - bline))
For j = 1 To 6
sumx(j) = sumx(j) + t(ct)
sumy(j) = sumy(j) + ch(j, ct)
Nextj
ct = ct + 1
Loop Until (t(ct) - bline) > 0
For j = 1 To 6
avgx(j) = sumx(j) / (ct - 1)
avgy(j) = sumy(j) / (ct - 1)
Nextj
' Determine average voltage for end of curve
ct = Nl -30
Do ' last 30 values
For j = 1 To 6
sumxl(j) = sumxl(j) + t(ct)
sumyl(j) = sumyl(j) + ch(j, ct)
Nextj
ct = ct + 1
Loop Until (ct-Nl) = 0
For j = 1 To 6
avgxl(j) = sumxl(j) / 30
avgyl(j) = sumyl(j) / 30
Nextj
ct = 1 'counter
' Calculate Slopes and Intercepts
For j = 1 To 6
Slope(j) = (avgy(j) - avgyl(j)) / (avgx(j) - avgxl(j))
slnt(j) = avgy(j) - (Slope(j) * avgx(j))
Nextj
1 MsgBox (Slope(6))
'MsgBox (slnt(6))
lblTemp.Caption = Format(avgy(6) * 100, "#0.0")
LblBLSlope.Caption = Format(Slope(6) * 36000, "O.OE-00")
LblBLInt. Caption = Format(slnt(6), "#0.000")
'MsgBox (Checkl)
If Checkl = 0 Then 'For drift correction
50
-------
'subtract baseline and keep variable name This routine is to compensate for drift in the
base
' line. To be used the sampling time should be atleast two times the first monemt of the
effluent
' transducer. Change the fort color is insuficient data to use.
For i = 0 To N
For j = 1 To 5
ch(j,i) = ch(j,i)-avgy(j)
1 Ifi= 1 Andj = 1 Then
' MsgBox (ch(j, i))
1 End If
Nextj
Nexti
Else
1 MsgBox ("Temp Cor")
1 MsgBox (ch(6, 1))
For i = 1 To N
For j = 1 To 5
chG, i) = ch(j, i) - (SlopeG) * t(i) + slnt(j))
Nextj
Nexti
1 MsgBox (ch(6, 1))
' MsgBox (slnt(6))
End If
' find time at maximum signal
' the time is calculated as max time - base line time - (pulse time/2)
' This will be one of the metrics evaluated to determine travel time
Dim jmax(6) As Integer 'sample number at max temperature
For i = 1 To 6
tempmax(i) = ch(i, 1)
tmax(i) = t(l)
Next i
For i = 1 To 6
Forj =2ToN
If (ch(i, j) - tempmax(i) > 0) Then
tempmax(i) = ch(i, j)
tmax(i) = t(j)
jmax(i)=j
End If
Nextj
Nexti
51
-------
Ibltmaxl.Caption = Format(tmax(l) - bline, "#0.0")
lbltmax2.Caption = Format(tmax(2) - bline, "#0.0")
Ibltmax4.Caption = Format(tmax(4) - bline, "#0.0")
IbltmaxS.Caption = Format(tmax(5) - bline, "#0.0")
'Begin Moment analysis - Data will be extrapolated based on a lograthmetic
' function based on a sum of squares fit to the data between 80% of tempmax
' and next occurance of <= 0.002 VDC or the end of file which ever comes first
' first determine sample number correlated to 80%of tempmax to perform moment
analysis
' there should be only one peak and it should occure in the first 40% of the data set
' if peak temperature occures after the 40% of the data points do not make moment
analysis
Dim jbegin(5) As Integer 'jbegin is the sample number for the first sample to use in
generating fit
Dim jend(5) As Integer 'jend is the last sample to use
If ((0.4 * Nl) - jmax(l) < 0) And ((0.4 * Nl) - jmax(2) < 0) And ((0.4 * Nl) -
jmax(3) < 0) _
And ((0.4 * Nl) - jmax(4) < 0) And ((0.4 * Nl) - jmax(5) < 0) Then
MsgBox (j)
MsgBox (Nl)
MsgBox ((0.4 *Nl)-j)
IblMoml. Caption = Format("N.A.")
lblMom2. Caption = Format("N.A.")
lblMom4. Caption = Format("N.A.")
IblMomS. Caption = Format("N.A.")
GoTo 200
End If
For i = 1 To 5
j=jmax(i)
Do Until ((tempmax(i) * 0.8) - ch(i, j) > 0) 'determine sample number for first sample
' to use in enterpolating and extrapolating
' moments 80% of peak was qualitatively
' estimated as being sufficient to be away
' from the curvature at the peek
(j -N = 0)Then
jbegin(i)=j
jend(i)=j
GoTo 90
End If
jend(i)=jend(i)+ 1
Loop
52
-------
jbegin(i)=j
j=j + l
If j >= Nl Then GoTo 89
Do Until (0.002 - ch(i, j) > 0) 'determine last sample number considered valid for
'interpolating or extrapolating moments noise was measured
'at +- 2 mv. data has already been adjusted for baseline
'offset
If j >= Nl Then GoTo 89
Loop
89 jend(i)=j
90 Next i
'MsgBox (t(jmax(l)))
'MsgBox (jbegin(l))
'MsgBox (jend(l))
' Fit log transformed data to a streight line first determine if there is enough data
' there should be a minimum of minobs observations to calculate the extrapolation
function
Dim sumxx(5)
Dim sumyy(5)
Dim sumxy(5)
Dim avgxx(5)
Dim avgyy(5)
Dim Mslope(5) As Double
Dim Mint(5) As Double
Dim minobs As Integer 'minimum number of observations between jbegin(i) and jend(i)
'to permit moment analysis 20 is considered minimm
Dim minsignal As Single 'minimum peak signal in volts for analysis
minobs = 20
minsignal = 0.001
For i = 1 To 5 'initialize values
sumxx(i) = 0
sumyy(i) = 0
sumxy(i) = 0
avgxx(i) = 0
avgyy(i) = 0
Mslope(i) = 0
Mint(i) = 0
Nexti
For i = 1 To 5
If ((jend(i) - jbegin(i) - minobs) > 0) And (tempmax(i) - minsignal > 0) Then
Forj = jbegin(i) To jend(i) - 1
53
-------
avgxx(i) = avgxx(i) + t(j)
If ch(i, j) < 0 Then ch(i, j) = 1E-50
avgyy(i) = avgyy(i) + Log(ch(i, j))
Nextj
avgxx(i) = avgxx(i) / (jend(i) - jbegin(i))
avgyy(i) = avgyy(i) / (jend(i) - jbegin(i))
Else
End If
Next i
For i = 1 To 5
If ((jend(i) - jbegin(i) - minobs) > 0) And (tempmax(i) - minsignal > 0) Then
Forj = jbegin(i) To jend(i) - 1
If ch(i, j) < 0 Then ch(i, j) = 1E-50
sumxx(i) = sumxx(i) + (t(j) - avgxx(i)) A 2
sumyy(i) = sumyy(i) + (Log(ch(i, j)) - avgyy(i)) A 2
sumxy(i) = sumxy(i) + (t(j) - avgxx(i)) * (Log(ch(i, j)) - avgyy(i))
Nextj
Mslope(i) = sumxy(i) / sumxx(i)
Mint(i) = avgyy(i) - (Mslope(i) * avgxx(i))
Else
End If
Nexti
'Calculate zero and first moments
Dim Momzero(5)
Dim Mom 1(5)
Dim cl, c2
Dim Tl, T2
Dim timecor 'time correction such that zero time is at the end of base line measurements
Dim j start 'sample number to begin moment calculations this is to remove switching
noise
timecor = bline
For i = 1 To 5
If ((jend(i) - jbegin(i) - minobs) < 0) Or (tempmax(i) - minsignal < 0 Or Mslope(i) > 0)
Then
Moml(i) = "N.A."
' GoTo 300
Else
Momzero(i) = 0
Moml(i) = 0
ch(i, 0) = 0
t(0) = 0
Start moment analysis beginning with sample after heat-pulse
Correct time scale to begin halfway through the heat-pulse in moment calculations
54
-------
' First step is to determine sample number after end of heat-pulse
j = l
Do Until (t(j) - bline - heater) > 0
j=j + l
Loop
j start = j - 1
For j = j start To jbegin(i) 'first part of curve
Momzero(i) = Momzero(i) + ((t(j) - t(j - 1)) * (ch(i, j) + ch(i, j - 1))) / 2
Moml(i) = Moml(i) + (1 / 6) * ((ch(i, j - 1) * (t(j - 1) - timecor) + ch(i, j) * _
(tG) - timecor)) * (t(j) - t(j - 1)) + (ch(i, j - 1) _
+ ch(i, j)) * ((t(j) - timecor) * (t(j) - timecor) - (t(j - 1) - timecor) * _
(t(j - 1) - timecor)))
Nextj
' MsgBox (i)
' MsgBox (Momzero(i))
' MsgBox (Moml(i))
' MsgBox (ch(ij))
' MsgBox (Mslope(i))
' MsgBox (Mint(i))
'end of first part of curve begin analysis for the curve fit portion of the curve
cl =ch(i,j)
Tl=t(j)
j = l
Do Until (cl - 0.001 * tempmax(i) < 0)
T2 = Tl + 1
c2 = Exp(Mslope(i) * (T2) + Mint(i))
Momzero(i) = Momzero(i) + (cl + c2) / 2
Moml(i) = Moml(i) + (1 / 6) * ((cl * (Tl - timecor) + c2 * (T2 - timecor)) * _
(T2-Tl) + _
(cl + c2) * ((T2 - timecor) * (T2 - timecor) - (Tl - timecor) * (Tl - timecor)))
Tl = T2
cl=c2
Loop
End If
3 00 Next i
'MsgBox (Moml(l))
If ((jend(l) - jbegin(l)) > minobs) And (tempmax(l) - minsignal > 0 And
IsNumeric(Moml(l))) Then
IblMoml.Caption = Format(Moml(l) /Momzero(l), "#0.0")
Else
IblMoml. Caption = Format("N.A.")
End If
55
-------
If ((jend(2) - jbegin(2)) > minobs) And (Not Not tempmax(2) - minsignal > 0 And
IsNumeric(Moml(2))) Then
lblMom2. Caption = Format(Moml(2) / Momzero(2), "#0.0")
Else
lblMom2.Caption = Format("N.A.")
End If
If ((jend(4) - jbegin(4)) > minobs) And (tempmax(4) - minsignal > 0 And
IsNumeric(Moml(4))) Then
lblMom4. Caption = Format(Moml(4) / Momzero(4), "#0.0")
Else
lblMom4. Caption = Format("N.A.")
End If
If ((jend(5) - jbegin(5)) > minobs) And (tempmax(5) - minsignal > 0 And
IsNumeric(Moml(5))) Then
IblMomS. Caption = Format(Moml(5) / Momzero(S), "#0.0")
Else
IblMomS. Caption = Format("N.A.")
End If
' generate labels for flow
DimCL
CL = 2.326
' Channel 0 Captions
If ((jend(l) - jbegin(l) > minobs) And (tempmax(l) - minsignal > 0) And _
(tmax(l) - bline - 1.5 * heater > 0) And IsNumeric(Moml(l))) Then
Moml(l) = (Moml(l) /Momzero(l))
QO = (clO(ii) - (c30(ii) * Moml(l))) / (Mom 1(1) + c20(ii))
QOmin = (clO(ii) - (c30(ii) * Moml(l) * (1 + CL * cvO(ii)))) / (Moml(l) * (1 + CL
* cvO(ii)) + c20(ii))
QOmax = (clO(ii) - (c30(ii) * Moml(l) * (1 - CL * cvO(ii)))) / (Moml(l) * (1 - CL *
cvO(ii)) + c20(ii))
'MsgBox (Moml(l))
'MsgBox (clO(ii))
IblQO.Caption = Format(QO, "#0.00")
IblQOmax. Caption = Format(QOmax, "#0.00")
IblQOmin.Caption = Format(QOmin, "#0.00")
Else
QO = "N.A."
QOmin="N.A."
QOmax="N.A."
IblQO.Caption = Format(QO)
IblQOmin.Caption = Format(QOmin)
IblQOmax. Caption = Format(QOmax)
End If
56
-------
Channel 1 Captions
If (jend(2) - jbegin(2) > minobs) And (tempmax(2) - minsignal > 0) And (tmax(2) -
bline- 1.5 * heater >0_
And IsNumeric(Moml(2))) Then
Mom 1(2) = (Mom 1(2) / Momzero(2))
Q 1 = (c 1 1 (ii) - c3 1 (ii) * Mom 1 (2)) / (Mom 1 (2) + c2 1 (ii))
Qlmin = (cll(ii) - c31(ii) * Moml(2) * (1 + CL * cvl(ii))) / (Moml(2) * (1 + CL *
Qlmax = (cll(ii) - c31(ii) * Moml(2) * (1 - CL * cvl(ii))) / (Moml(2) * (1 -
cvl(ii)) + CL*c21(ii))
IblQl. Caption = Format(Ql, "#0.000")
IblQl max. Caption = Format(Qlmax, "#0.000")
IblQlmin.Caption = Format(Qlmin, "#0.0")
Else
Q1 = "N.A."
Qlmin = "N.A."
Qlmax ="N.A."
IblQl. Caption = Format(Ql)
IblQlmin.Caption = Format(Qlmin)
IblQl max. Caption = Format(Qlmax)
End If
Channel 3 Captions
If (jend(4) - jbegin(4) > minobs) And (tempmax(4) - minsignal > 0) And _
(tmax(4) - bline - 2 * heater > 0 And IsNumeric(Moml(4))) Then
Mom 1(4) = (Mom 1(4) / Momzero(4))
Q3 = (c!3(ii) - c33(ii) * Mom 1(4)) / (Mom 1(4) + c23(ii))
Q3min = (c!3(ii) - c33(ii) * Moml(4) * (1 + CL * cv3(ii))) / (Moml(4) * (1 + CL *
cv3(ii)) + c23(ii))
Q3max = (c!3(ii) - c33(ii) * Moml(4) * (1 - CL * cv3(ii))) / (Moml(4) * (1 - CL *
cv3(ii)) + c23(ii))
'MsgBox (c!3(ii))
'MsgBox (c33(ii))
'MsgBox (c23(ii))
1WQ3. Caption = Format(Q3, "#0.000")
lblQ3max.Caption = Format(Q3max, "#0.000")
lblQ3min.Caption = Format(Q3min, "#0.000")
Else
Q3 = "N.A."
Q3min="N.A."
Q3max="N.A."
lblQ3. Caption = Format(Q3)
lblQ3min. Caption = Format(Q3min)
lblQ3max. Caption = Format(Q3max)
End If
Channel 4 Captions
If (jend(5) - jbegin(5) > minobs) And (tempmax(5) - minsignal > 0) And _
57
-------
(tmax(5) - bline - 1.5 * heater > 0 And IsNumeric(Moml(5))) Then
Mom 1(5) = (Mom 1(5) / Momzero(5))
Q4 = (c!4(ii) - c34(ii) * Moml(5)) / (Moml(5) + c24(ii))
Q4min = (c!4(ii) - c34(ii) * Moml(5) * (1 + CL * cv4(ii))) / (Moml(5) * (1 + CL *
cv4(ii)) + c24(ii))
Q4max = (c!4(ii) - c34(ii) * Moml(5) * (1 - CL * cv4(ii))) / (Moml(5) * (1 - CL *
cv4(ii)) + c24(ii))
'MsgBox (Q4)
lblQ4.Caption = Format(Q4, "#0.000")
lblQ4max.Caption = Format(Q4max, "#0.000")
lblQ4min.Caption = Format(Q4min, "#0.000")
Else
Q4 = "N.A."
Q4min="N.A."
Q4max="N.A."
lblQ4.Caption = Format(Q4)
lblQ4min.Caption = Format(Q4min)
lblQ4max. Caption = Format(Q4max)
End If
' Generate Darcy velocity tables
If DomeDia > 0.1 Then 'Data exists to calculate flux
IfQO<>"N.A."Then
VO = QO * 60 * 24 / (3.1415 * DomeDia * DomeDia / 4)
IblVO.Caption = Format(VO, "#0.0")
Else
VO = "N.A."
IblVO.Caption = Format(VO)
End If
IfQK>"N.A."Then
VI =Q1 * 60* 247 (3.1415 * DomeDia * DomeDia / 4)
IblVl.Caption = Format(Vl, "#0.0")
Else
VI = "N.A."
IblVl.Caption = Format(Vl)
End If
IfQ3<>"N.A."Then
V3=Q3 * 60* 247 (3.1415 * DomeDia * DomeDia / 4)
lblV3.Caption = Format(V3, "#0.0")
Else
V3 = "N.A."
lblV3.Caption = Format(V3)
End If
IfQ4<>"N.A."Then
V4 = Q4 * 60 * 24 / (3.1415 * DomeDia * DomeDia / 4)
58
-------
lblV4.Caption = Format(V4, "#0.0")
Else
V4 = "N.A."
lblV4.Caption = Format(V4)
End If
Else
VO = "N.A."
VI = "N.A."
V3 = "N.A."
V4 = "N.A."
IblVO.Caption = Format(VO)
IblVl.Caption = Format(Vl)
lblV3.Caption = Format(V3)
lblV4.Caption = Format(V4)
End If
' set up arrays for ploting
200 Dim GraphQ As Single
Dim x As Integer
Dim myXarrayQ As Double
Dim my YArrayOQ As Double
Dim myYArraylQ As Double
Dim my YArray2() As Double
Dim myYArraySQ As Double
Dim my YArray4() As Double
Dim my YArraySQ As Double
' Dim myXarray5(l) As Double
' Dim myYArray5(l) As Double
ReDim myXarray(N)
ReDim myYArrayO(N)
ReDim myYArrayl(N)
ReDim myYArray2(N)
ReDim myYArray3(N)
ReDim myYArray4(N)
ReDim myYArray5(N)
'Generate some x y data.
myXarray(O) = t(l) - bline
myYArrayO(O) = ch(l, 1) * 10
myYArrayl(O) = ch(2, 1) * 10
59
-------
myYArray2(0) = ch(3, 1) * 10
myYArray3(0) = ch(4, 1) * 10
myYArray4(0) = ch(5, 1) * 10
myYArrayS(O) = ch(6, 1) * 100
1 myYArray5(0) = 0
1 myYArray5(l) = 0.03
' myXarray5(0) = 0
' myXarray5(l) = heater
' MsgBox n
For x = 1 To N - 1
myXarray(x) = t(x) - bline Value for X-axis
myYArrayO(x) = ch(l, x) * 10 Value for Y-axis
myYArrayl(x) = ch(2, x) * 10
myYArray2(x) = ch(3, x) * 10
myYArray3(x) = ch(4, x) * 10
myYArray4(x) = ch(5, x) * 10
myYArray5(x) = ch(6, x) * 100
Nextx
With TChartl
.AddSeries scPoint
.Series(0).AddArray UBound(myYArrayO), myYArrayOQ, myXarrayQ
.Series(0).XValues. Temp Value = True
.Series(l).AddArray UBound(myYArrayl), myYArraylQ, myXarrayQ
.Series(2).AddArray UBound(myYArray2), myYArray2(), myXarrayQ
.Series(3).AddArray UBound(myYArray3), myYArray3(), myXarrayQ
.Series(4).AddArray UBound(myYArray4), myYArray4(), myXarrayQ
.Series(5).AddArray UBound(myYArray5), myYArraySQ, myXarrayQ
End With
Screen.MousePointer = 1
' write output file with processed data
ChDir Dirl
Dim fname2 ' file name for output data
fname2 = Format$(timel, "mmm dd hh nn ss") + "a.CSV"
Open fname2 For Output As #8
Write #8, Format("baseline"), Format("heater"), Format("Ref Resistor"),
Format("Dome D"), Format("site")
Write #8, Format(bline, "#0.0"), _
Format(heater, "#0.0"), Format(CalResistor, "#0.0000"), _
Format(DomeDia, "#0.0000"), Format(site)
60
-------
Write #8, Format("Peak arrival time in seconds")
Write #8, Format(tmax(l), "#0.0"), Format(tmax(2), "#0.0"), Format(tmax(3),
"#0.0"), _
Format(tmax(4), "#0.0"), Format(tmax(5), "#0.0")
Write #8, Format("Maximum differential temperature or reference temperature c")
Write #8, Format("Ch 0"), Format("Ch 1"), Format("Ch 2"), Format("Ch 3"),
Format("Ch 4"), _
Format("RefTemp")
Write #8, Format(tempmax(l) * 10, "#0.0"), Format(tempmax(2) * 10, "#0.0"),
Format(tempmax(3) * 10, _
"#0.0"), Format(tempmax(4) * 10, "#0.0"), Format(tempmax(5) * 10, "#0.0"),
Format(avgy(6) * 100, "#0.0")
Write #8, Format("Normalized First Moments Calculated beginning at the end of the
base line samples")
If ((jend(l) - jbegin(l)) > minobs) And (tempmax(l) - minsignal > 0 And _
Mom 1(1) <> 0 And IsNumeric(Moml(l))) Then
Write #8, Format("Ch 0"), Format(Moml(l), "#0.0")
Else
Write #8, Format("Ch 0"), Format("N.A.")
End If
If ((jend(2) - jbegin(2)) > minobs) And (tempmax(2) - minsignal > 0 And _
Mom 1(2) <> 0 And IsNumeric(Moml(2))) Then
Write #8, Format("Ch 1"), Format(Moml(2), "#0.0")
Else
Write #8, Format("Ch 1"), Format("N.A.")
End If
If ((jend(3) - jbegin(3)) > minobs) And (tempmax(3) - minsignal > 0 And _
Moml(3) <> 0 And IsNumeric(Moml(2))) Then
Write #8, Format("Ch 2"), Format(Moml(3), "#0.0")
Else
Write #8, Format("Ch 2"), Format("N.A.")
End If
If ((jend(4) - jbegin(4)) > minobs) And (tempmax(4) - minsignal > 0 And _
Mom 1(4) <> 0 And IsNumeric(Moml(4))) Then
Write #8, Format("Ch 3"), Format(Moml(4), "#0.0")
Else
Write #8, Format("Ch 3"), Format("N.A.")
End If
If ((jend(5) - jbegin(5)) > minobs) And (tempmax(5) - minsignal > 0 And _
Mom 1(5) <> 0 And IsNumeric(Moml(5))) Then
Write #8, Format("Ch 4"), Format(Moml(5), "#0.0")
Else
Write #8, Format("Ch 4"), Format("N.A.")
End If
Close
Refresh
61
-------
DoEvents
GoTo 45
44 Unload Me
Screen.MousePointer = 1
45 End Sub
Private Sub TeeCommander2_OnEditedChart()
End Sub
Private Sub Mom4_Click()
End Sub
62
-------
Appendix D Sample raw data file
This raw data is one sample run from the calibration data set for the flow meter
used at Lake Hartwell. After the first row that describes the data being collected the first
column is the date time the sample was collected, the second column it elapsed time, the
next 5 columns are differential temperature and the last column with data is the absolute
temperature.
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
•
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
:15:
:15:
:15:
:15:
:15:
:15:
:15:
:15:
:15:
:15:
•16:
•16:
:16:
:16:
:16:
:16:
•16:
•16:
:16:
:55 AM","
:55 AM","
:56 AM","
:56 AM","
:57AM","
:57AM","
:58 AM","
:58 AM","
:59 AM","
:59 AM","
:00 AM","
:01 AM","
:01 AM","
:02 AM","
:02 AM","
:03 AM","
:03 AM","
:04 AM","
:04 AM","
: 16:05 AM","
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:16:
:06 AM","
:06 AM","
:07 AM","
:07 AM","
:08 AM","
:08 AM","
:09 AM","
:09 AM","
:10AM","
:10AM","
:11 AM","
:12 AM","
:12 AM","
:13 AM","
:13 AM","
:14AM","
:14AM","
:15 AM","
20.0", "10. 0","99996.9482","0. 1969", "Purdue Calibration pump 15 6 ml/min
0.00000","-0.13058","-0.13232","-0.09803","-0.12581","-0.13805","0.24682","",'
0.55078","-0.13067","-0.13236","-0.09801","-0.12579","-0.13803","0.24698","",'
1.11133","-0.13054","-0.13229","-0.09806","-0.12578","-0.13804","0.24772","",'
1.66211","-0.13053","-0.13229","-0.09799","-0.12590","-0.13794","0.24880","",'
2.21289","-0.13052","-0.13237","-0.09803","-0.12582","-0.13806","0.24664","",'
2.76367","-0. 13059","-0. 13239","-0.09799","-0. 12574", "-0. 13806","0.24853","",'
3.31445'',"-0.13053'',"-0.13224'',"-0.09799'',"-0.12572'',"-0.13805","0.24754","",'
3.86523","-0.13056","-0.13227","-0.09803","-0.12578","-0.13808","0.24922","",'
4.41602","-0.13058","-0.13231","-0.09798","-0.12575","-0.13805","0.24859","",'
4.96680","-0.13057","-0.13232","-0.09802","-0.12578","-0.13811","0.24807","",'
5.51758","-0.13050","-0.13232","-0.09797","-0.12575","-0.13801","0.24752","",'
6.06836","-0.13062","-0.13236","-0.09801","-0.12578","-0.13814","0.24856","",'
6.61914","-0.13055","-0.13229","-0.09803","-0.12570","-0.13816","0.24842","",'
7.16992","-0.13057","-0.13234","-0.09801","-0.12583","-0.13802","0.24701","",'
7.72070","-0. 13049","-0. 13238","-0.09792","-0. 12579", "-0. 13805","0.24820","",'
8.27148","-0.13060","-0.13229","-0.09808","-0.12587","-0.13801","0.24844","V
8.82227","-0.13061","-0.13242","-0.09800","-0.12575","-0.13804","0.24648","",'
9.37305","-0.13066","-0.13222","-0.09797","-0.12567","-0.13812","0.24821","",'
9.92383","-0.13061","-0.13224","-0.09802","-0.12577","-0.13809","0.24707","",'
10.47461","
11.02539","
11.57617","
12.12695","
12.68750","
13.23828","
13.78906","
14.33984","
14.89063","
15.44141","
15.99219","
16.54297","
17.09375","
17.64453","
18.19531","
18.74609","
19.29688","
19.84766","
-0.13054","
-0.13053","
-0.13054","
-0.13051","
-0.13059","
-0.13051","
-0.13064","
-0.13053","
-0.13058","
-0.13058","
-0.13052","
-0.13050","
-0.13055","
-0.13058","
-0.13060","
-0.13051","
-0.13059","
-0.13059","
-0.13239","
-0.13227","
-0.13229","
-0.13231","
-0.13230","
-0.13241","
-0.13230","
-0.13231","
-0.13235","
-0.13242","
-0.13228","
-0.13229","
-0.13229","
-0.13243","
-0.13221","
-0.13233","
-0.13239","
-0.13231","
-0.09798","
-0.09799","
-0.09802","
-0.09804","
-0.09799","
-0.09802","
-0.09800","
-0.09804","
-0.09804","
-0.09801","
-0.09798","
-0.09804","
-0.09802","
-0.09794","
-0.09800","
-0.09799","
-0.09798","
-0.09796","
-0.12579","
-0.12576","
-0.12578","
-0.12578","
-0.12573","
-0.12583","
-0.12580","
-0.12570","
-0.12576","
-0.12574","
-0.12575","
-0.12575","
-0.12591","
-0.12574","
-0.12580","
-0.12572","
-0.12565","
-0.12580","
-0.13799","
-0.13804","
-0.13803","
-0.13798","
-0.13807","
-0.13800","
-0.13804","
-0.13802","
-0.13805","
-0.13804","
-0.13803","
-0.13803","
-0.13798","
-0.13812","
-0.13795","
-0.13809","
-0.13810","
-0.13809","
0.24881",""
0.24713",""
0.24707",""
0.24700",""
0.24856",""
0.24851",""
0.24681",""
0.24713",""
0.24748",""
0.24700",""
0.24683",""
0.24926",""
0.24861",""
0.24767",""
0.24823",""
0.24745",""
0.24944",""
0.24720",""
•
. (truncated)
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
"5/26/2006
5
5
5
5
5
5
5
5
5
5
5
5
5
•46:
:46:
•46:
•46:
:46:
:46:
:46:
•46:
•46:
•46:
:46:
:46:
:46:
:17AM","
:17AM","
:18AM","
:19AM","
:19 AM","
:20 AM","
:20 AM","
:21 AM","
:21 AM","
:22 AM","
:22 AM","
:23 AM","
:24 AM","
1821.87900
1822.43000
1822.98000
1823.53100
1824.08200
1824.63300
1825.18400
1825.73400
1826.28500
1826.83600
1827.38700
1827.93800
1828.48800
","-0.13064
","-0.13068
","-0.13073
","-0.13067
","-0.13074
","-0.13070
","-0.13070
","-0.13072
","-0.13067
","-0.13067
","-0.13068
","-0.13070
","-0.13066
","-0.13226
","-0.13232
","-0.13239
","-0.13246
","-0.13245
","-0.13248
","-0.13232
","-0.13236
","-0.13244
","-0.13232
","-0.13242
","-0.13238
","-0.13235
","-0.09814
","-0.09811
","-0.09812
","-0.09816
","-0.09815
","-0.09815
","-0.09816
","-0.09813
","-0.09812
","-0.09814
","-0.09816
","-0.09815
","-0.09812
","-0.12604
","-0.12606
","-0.12609
","-0.12616
","-0.12611
","-0.12607
","-0.12600
","-0.12600
","-0.12618
","-0.12608
","-0.12611
","-0.12617
","-0.12604
","-0.13821
","-0.13823
","-0.13826
","-0.13828
","-0.13816
","-0.13822
","-0.13827
","-0.13816
","-0.13814
","-0.13816
","-0.13817
","-0.13807
","-0.13818
","0.24751"
","0.24944"
","0.24939"
","0.24951"
","0.24583"
","0.24917"
","0.24893"
","0.24669"
","0.24928"
","0.24819"
","0.24698"
","0.24694"
","0.24982"
63
-------
"5/26/2006 5:46:24 AM","1829.03900","-0.13071","-0.13236","-0.09814","-0.12618","-0.13809","0.24636"
Appendix E Results from Hammond, IN
* Signal Display
tSblien
t3my documents
£3 flux meter
^ Hammond IH
1.2
1.1
1 •
0.9
0.8
9i 0.7
1
u OS
!°-5
~ "»
^ 0.3
0.2
0.1
0
-0.1 -
-0.2
Apr 21 O75738.dat
Apr 21 083657.dat
Apr 21 Q91617.dat
Apr 21 10Q735.dat
Temperature Response
; ;
;.——
0 100
"**•. i.
*•"..».
"•••
EBB
15 Cmed for Thermal drill?
Display Sensor Serial Ho default
Site Hammond R2S
**•
~**-
200 300 400
Time (sec)
•* »!-•
~=*«
mmtft. ^.
2
§
OS
1
son son TOO
PeakArnval First Moment Flow rate Flow 38% Confidence Limits Darcy velocity
Time fsecl fsec] Iml fml/minl Minimum Maximum cm/davl
Heatef (2] Ch 0 ^[jg|j N A NA NA NA N A
^^^^9 26.7 CH1 |£[£J 407.3 2.937 2.8 3.437 1.3
^emp'sfape6' a2E'02 Ch 3 |^Q 34S-2 '3-782 -3-521 -4 °81 ~2-4
, Ph ' , 0.267 Ch 4 441 3 NA HA N.A. HA H.A.
Intercept
5| my documenls
fluxmeler
JHarnmondlH
Apr 2018 50 00 dat
Apr20192427.dat
Apr 20 19 41 3S.dat
Apr 21 08 36 57.dat
w Correct [or Thermal drift?
Display Sensor Serial No default
Site Hammond River 12 South Ho Addition
Temperature Response
Peak Arrival
Time [sec]
00 SOO 1.000
Time (sec)
First Moment Flow rate
[seel
Heater (2) Ch 0
'T'emp-'sC' -1.SE-01 <*3
, (C'hr| 0.177 Ch4
Intercept
NA.
NA.
[ml Iml/minl
NA
NA
-3307
NA.
NA.
N.A.
-3.081
N.A.
NA.
NA.
-3565
NA.
Flow 38S Confidence Limits Darcy velocity
Minimum Maximum (cm/dayl
N.A.
-2.1
N.A.
64
-------
Correct for Thermal drift?
Display | Sensor Serial No default
Site Hammond River T2 South NoAdditio
Temperature Response
300 1 ,000
Time (sec)
Fksl Moment
[sec]
N.A.
4992
Flow rate
[ml ImlVmit
N.A.
-1.242
-1.748
Minimum
N.A.
-1.155
-1.682
M ax imum
N.A.
-1.340
-1.829
Darcy velocity
[cm/day]
-0.8
-1.1
(4)
* Signal Display 1-lfnfSl
£a
p
~
1.1
1
0.9-
0.8
0.7
1,8
^ 0.5
e
OJ
^ 0.3
02
0.1
0
-0.1
blien
my documents
Iflwmete
J Hammond IN
_g^^^^^^^^^^^| v
Apr 21 07 57 3B.dat ~
Apt21083E57.dat
Api21100735.dat
Apr21104654.dat "
Temperaiure Response
IS Correct for Thermal drift?
Display Sensor Serial No default
9ite Hammond River T2 South NoAddttion
f?
X
•_
0 200 40
Pea
Tim
Heater [2) ChO |
BfempSbpeel -3-8E-°3 Ch3 1
(C/hrj 016g Ch4
Intercept
0 61
Arrival
e(sec]
isa
gjMj
oa
0 800 1 ,000
Time (sec)
*=5=SS^—
1 ,200 1 ,4
First Moment Flow rate Flow 98^ C
(seel [ml [ml/mini MinimUm
NA NA N.A
6031.8 -0.428 -0.4
8703 -1227 -1.141
•» — z
^S" •
0
o|
00 1 ,600 1 ,800
onfidence Limits Darcy velocity
Maximum Icm/davl
NA. NA.
-0.405 -0.3
-1.323 -0.8
-4.681 -2.7
c.ie
65
-------
Apt 21 07 57 38.dat
i.pi;i 083657.dat
Apr 21 091617.dat
JtBhUilideEBSa
Apr 21 1046 cj.| dai
[^ Correct [OF Thermal drift?
Display Sensor Serial No default
Site Hammond River T2 South 5 ml/min Addition
Temperature Response
Peak Arrival
Time [seel
00 800 1 ,000
Time (sec)
First Moment
Healel (21 Ch 0
17.0 Chi
[seel
N.A.
240.7
342.4
Flow rate
[ml ImlVmit
N.A.
•6009
-6.974
Flow 98£ Confidence Limits
Minimum Maximum
N.A. NA
-5.572 -6.517
-6.348 -7.767
larcy velocity
[cm/davl
N.A.
NA
-3.8
-4.4
Display | Sensor Serial No default
| v | Site Hammond River T2 South 5 ml^min Addition
Temperaiure Response
800 1,000
Time (soc)
First Moment
(sec)
N.A.
NA
293.9
420.0
Flow rate
[ml (ml/minl
N.A.
NA
-4.645
-5.683
Flow 98% Confidence Limits
Minimum Maximum
NA NA
-4.319
-5.210
-5.022
-6.277
> arcy velocity
Icm/davl
NA.
NA
-2.3
-3.6
66
-------
(7)
* Signal Display
A Apr 21 08 36 57.dat
Apr 21 091617.dat
Apr 21 10 07 35 dat
Apr 21 11] 46 54 dar
v Correct for Thermal drift?
Display Sensor Serial No default
I I
Site Hammond River T2 South 5 ml/min Addition
Temperature Response
1.7E-01 Ch3
0.203 CM 2313
NA
N.A.
287.5
416.1
600
Time (sec)
Itlt Flow fate
[ml [ml/minl
NA.
N.A.
-4.775
-5.734
Fta 985: C
N.A.
N.A.
-4.439
-5.255
NA.
N.A.
-5.164
-6.336
Darcy velocity
(cin^dwl
N.A.
N.A.
-3.0
-3.6
67
-------
Appendix F Results from Lake Hartwell, SC
Mau 1)31531 ll'Jdal
May 03 15 53 03.dat
May031E1520.dat
May 031E 41 55 dal
I? Correct for Thermal drift?
Display Sensor Serial No default
Site Lake Hartwell, DB
Temperature Response
- 0.1
; 0
-0.1
-0.2-
-0.3
-0.4-
-H-
Heater (2]
Peak Arrival
Time [sec]
24.3 CM
2.2E-01 Ch3
0.249 Ch4 1779.3
800 1,000
Time (sec)
First Moment
181.5
122.0
N.A.
N.A.
Flow rale
[ml [ml/minl
15.68
1,200 1,400 1,600
Row r&% Confidence Limits
N.A.
N.A.
Minimum
14.18
N.A.
N.A.
Maximum
17.56
N.A.
N.A.
Darcy velocity
(cm/dawl
3.9
8.9
N.A.
N.A.
* Signal Display CTfnlfX]
1
Q.
1
i
t3 blien
(^ my documents
Q| flux meter
May O3145510.dat A I*
Correct for Thermal diift? j
Mfy'os'VK'o'g'da™ I DlsP|aV I Sensor Serial No default
May031El520.dat L '
May031E4155.dat v site Lake Harlwell, OB
Temperature Response
0-
-0.1 -
:ct
Jlfl^
V...w"
^
%«i
-.*!, ^
0 1 00 200
Peak
^-^
"
.'.v-.1-
- -P-
^' *
**
,^^,^
j.
S
of
1
300 400 500 600 700 800 900 1 ,000 1,100
Time (sec)
Arrival First Moment Flow rate Flow '38=; Lunhdpnce Limits Darcy velocity
Time [seel [seel [ml [ml/mini Minimum Maximum (cm/davl
Heater [2] Ch 0 |
•
Te
^^B 26.6 Chi |
Aslope* -1-5E-Q2 Ch 3 |
(!:yhr) 0 266 Ch 4
211.5 12.63 11.49 14.02 7.9
jjjjj 190.2 7.914 7.5 10.681 5.0
JgJ N.A. N.A. N.A. N.A. N.A.
987 NA NA NA N.A N.A
Intercept
68
-------
(3)
* Signal Display |3@®
Rel Temp (C deg)
o o ooooooooo
blien
my documents
I Flux meter
^LakeHartraell
JES^^^M-'
May03145510.dat
MayQ316152Q.dat
May031E4155.dat v
Temperature Response
^
.: m
; \V_
-•-^j^,
f
•
j? Correct for Thermal drift?
Display Sensor Serial No default
Site Lake HaKwel, 08
1 ^y 1^
-*— w
0 200
400 600
Time (sec)
,-..-^r. '—_-—
-*t
-^,;
Q
Temperature (C)
SOO 1 ,000 1 ,200
ose
PeakArrival FirslMoment Florarae Flora 985i Confidence Limits Dacy velocity
Timefsecl Isecl [m! [ml/mini Minimum Maximum cm^dav]
Heater (2)
73.9
55.3
N.A.
N.A.
N.A.
N.A.
103.19
47.1
N.A.
N.A.
N.A.
N.A.
92.2
322
N.A.
N.A.
Rel Temp (C deg)
oo ooooooooo
Temperature Response
laS
.-f
»
SS^!Si>t%(
W.--*4-*^.
•^
- - mttty**
0 200 400 600 800 1,C
Time (sec)
"PeakAnival
Time (seel
Heater (2) Ch 0 g|m|
BT"mp™opeB' 1.1E-02 ^ ^
(C/hr) 0262 ch< g176
First Moment
1089.2
SB5.0
N.A.
2361.9
Flora rale Flora S8% Confide
[ml fml/nin] Minimum f
2.13 201
0.9G4 0.9
N.A. N.A.
-1739 -1674
^»j
00 1 .20
icf Limit: Da
aximum
2.29
1.05S
N.A.
-1.819
ij
o|
I
0
- >,' 'vvluc il1,1
=m/davl
1.3
0.6
N.A.
-11
69
-------
(5)
* Signal Display Qfnjfx]
S?
u
E
$
i^b|ien ^
Q) my documents
£3 flux meter
^p^akej^^^^_
Maj.03145510.dat
May0315310S.dat
May0315530S.dat
'jjLlll3iii1|6|\Vl°i|dM^ ^ ••
Temperature Response
8-
-0.2
^
:*X
tifflc
.j^.TX- rfM**SSfc*. ,.y
T""'"" "^ V"
v \- :- \- ---
^? Correct for Thermal drift?
I Display I Sensor Serial No default Clo!e
Site Lake Hailwel, 08
. ,--•
-V
-;-.—•
0 200 400 600 300 1,000
Time (sec)
Peak Arrival First Moment Flow ra
* '
r_j M*X jj
-
V*
1 ,200 1 ,400 1 ,500 1 ,8C
ff
i
°I
1
0
a Flom 98X Confidence Limits Darcy velocity
Time fsecl (seel [m! fmVmm} Minimum Maximum cmydavl
Healer (2) Ch 0 gjgj 1 70.8 1 7.19 15.48 1 8.32 1 0.8
BIH9 2S1 CM EUJ 1459 11-154 10G 17-110 7-°
Bas
nn'sLir -5.3E-02 Ch 3 H&HI N.A. N.A-
, fC/hr) 0.261 Ch4 548.7 N.A. N.A.
Intercept
N.A. N.A. N.A.
N.A. N.A. N.A.
Rel Temp (C deg)
a
_
1 •
0.9-
0.8-
0.7-
0.6
0.5-
0.4-
0.2-
0.1
0
0.1 -
0.2-
blien
my documents
i flux meter
JLakeHartwel
'-
A
?\
|fei
> -"—-"—
rV«
-------
(7)
* Signal Display Qfnjfx]
tablien A. May 03 16 15 20 dat ~ V Correct [or Thermal drift?
e3 my documents Maji03164155.dat I I . Close
Q|tomBter MVlVlVlVlM I I '
^yJQ|£j^j^^_ -fM7y04085533.dat v Site Lake Harlwel, 08
Temperature Response
0.8
ij
Lt
^*«E/*W-»
; W'i*^as**"^3aZ
i . * i "^^S!::::i*a»»^-, :
0 50 100 150 200
8
oi
1
Time (sec)
Peak Arrival First Moment Flora ra e Flow 3B% Confidence Limits Daley velocity
Time (seel Isecl (ml (ml/mini Minimum Maximum cmydavl
Heater (2) ChO fgf 113.1 32.25 28.12 37.82 20.3
BHHB 223 Ch1 E£l 837 2t505 23° 86739 m"
Bfemp'sfopee' 1 2EO° ch 3 t™* N.A. N.A. N.A. N.A. NA
, (C*rI 0.223 Ch 4 78.9 N.A. N.A. N.A. N.A. N.A.
Intercept
t^blien
t3 mv documents
Ql flux meter
'ESLake Hailwell
1
0.9
0.8
0.7
|.6
I"
0.2
0.1
0
-0.1
-
JW
£^%
V
^
V
•-
May 03 16 15 20 dat
May 03 16 41 55.dat
May03171347.dat
Temperature Response
^- ^i^S=^--_
IS Correct for Thermal drift?
Display Sensor Serial No default
I '
Site Lake Harlwel, 08
0 100 200 300 400 500
Time (sec)
**~- —
6t
Peak Arrival First M oment Flow rate
Time [sec] Isecl [m] (ml/minl
Heater (2) Ch 0 [QQ 1 53.5 1 3.1 2
^empsfope^ 47E-°1 Ch 3 U3 43'° 1 36'1 20
, *C/hr] 0234 Ch4 785.8 NA NA
Intercept
*—_\
o
Temperature (C)
0 700 800 300
Flow 38^ Confidence Limits Da
Minimum Maximum
17.15 21.61
83 12.131
182.848 105.172
N.A. N.A.
cm/davl
12.0
5.5
85.6
NA
ose
71
-------
(9)
* Signal Display Qfnjfx]
£3 Hen
H^ my documents
t3 flu* meter
V_LEZ^^^H
§ 0,
u
1 as
KL
°
Ma,p4085533da, ~
Has 04 09 29 56.dat
May 04 09 44 55.dat
May 04 09 59 20.dat »
& Correct for Thermal drift?
I Display I Sensor Serial No default Clo!e
Site Lake Hailwel, 08
Temperature Response
/
|
'
-------
(11)
* Signal Display Qfnjfx]
e
t
1.1
1
0.9
0.8
2 0.6
|0.5
T 04
0.3
0.2
0.1
0-
Iblien
j| my documents
3 flux meter
^LakeHartwell
JH^^^^^H <
May04085533.dat
May04091457.dat
May 04 03 59 20 dat' V
li? Correct for Thermal drift?
1 Display 1 Sensor Serial No default Clo!e
Site Lake Harlwel, OB
Temperature Response
-.
/
.
:;— -
%-v^
/^
f...
1" I ,
••-.
0 50 100
^r...__r — : : : ^»__
M — • : ^">"'^=-~
- , ': : ': : ; : "' :
150 200 250 300 350 400 450 5C
Time (sec)
Peak Arrival First Moment Flora ra e Flow 3B% Confid!
Time fsecl Isecl [ml (ml/rninl Minimum
Heater |2) ChO J2JJ 482.8 4.72 4.38
BJBH9 251 Ch1 E3 3762 3-330 a2
eT7mpUsTopeB' 4-2E-02 » 3 ^Q 547.9 -2.160 -2.014
, (C*rI 0.251 Ch 4 370.9 N.A. N A N A
Intercept
0 55
nee Limit
1 axir:-ium
5.11
3.856
-2.326
N.A.
o|
1
0 600
B Darcy velocity
cmydavl
3.0
21
-1 4
N.A.
(12)
* Signal Display 0[&®
Re! Temp (C deg)
_J
_j
1.1 •
1 -
0,3
0.8
0.7
0.8
0.5
0.4-
0.3
0.2-
0.1 •
0
lien
my documents
flux meter
LakeHaifrVeil
|^^^^^^^^| v
.*/>>
: f ^
I ^*
[171
May 04 OB 55 33.dat
May 04 09 14 57 dat
May 04 09 29 5E dat
Temperature Response
£
\
0 100
f/ Correct for Thermal diift?
Display Sensor Serial No default
I '
Site Lake Harlraell, OB
200 300 400
Time (sec)
o
Temperature (C)
500 600 700
ose
Peak Anival Fir:,i Moment Flow rate Flow 93% Lonfidence Limits Darcy velocity
Time fsecl Isec] [m! [ml/mini Minimum Maximum cm/dav]
Heater (2) Ch 0 ^Tjj] 120.9 31.27 27.33 36.57 19.7
HBBB 25-3 Ch 1 EO 89-° 22.279 20.9 64.722 1 4.0
^empSlope3' ^ 2E'02 Ch3 t^£l NA NA NA NA NA
fC/hf) 0253 Ch 4 1S4? 3iJ34 7 76g 7 044 _8692 _4g
73
-------
(13)
* Signal Display EM®
f.
1.1
1 •
0.9
08
m07
Q.
E 0.5-
«0.4
0.3
0.2
0.1
0
blien
j| my documents
3 flux meter
^LakeHartwell
_33££^^^^H v
/
! : S
May 04 10 33 20. dat
May0410575l.dat
May 04 11 0602.dat
May04112253.dat v
Temperature Response
^-^
—
0 100
i:
— __
j? Correct for Thermal drift?
Display 1 Sensor Serial No default
1 1
Site Lake HaKwel, OB
^r;:=:~' i •! "-^5
— — — . — . .
200 300 400
Time (sec)
— — m,* •
" ~~
: i
I
500 600 700
ose
Peak Arrival First Moment Flora ra e Flo™ 98X Confidence Limits Darcy velocity
Time fsecl Isecl fmlfml/min) Minimum Maximum cmyday]
Heater(2) ChO [|QJ 321.4 7.43 6.85 8.13 4.7
BTemp™ope!' 3-8E-°3 Ch 3 EH] NA N A N A N A NA
, (C*rI 0.252 Ch 4 53.4 N.A. N.A. N.A. N.A. N.A.
Intercept
(14)
a0-7
1
M.^rj41iW51.dat
May 04 1 1 06 Q2.dat
May 04 1 1 22 53.dat
M Correct for Thermal drift?
Display Sensoi Serial No default
' - '
Site Lake HaffAiell, OB
Temperature Respons
Fbw rate
[mHml/min! Minimum
6.35 6.42
500 600
Flow '&-, i:.,:.nfidence Limits
Maximum
7.59
Darcy velocity
cmyday]
74
-------
(15)
Ma.,.0.|101620.dat
Mar M 10 33 20 dat
May 04 11 06 02.dat
Maji04112253.dat
1^ Correct for Thermal drift?
Display 5emor Serial No default
Site Lake Harlwel, 08, S.S. 5 ml/min
Temperature Response
1.1
1 •
0.9
O.S
a07
oO
0.1
0
300
350
400
Flora B8% Confidence Limit
Minimum Maximum
14.61 18.15 10.2
Darcy velocity
[cm/daiil
12.E
N.A.
N.A.
22.566
N.A.
N.A.
8.4
N.A.
N.A.
(16)
* Signal Display 0[&®
t3 my documents
£j flux meter
tSLakeHartweil
KI&^^H
1.1
1
0.9
0.8
S07'
-8os
a.
|0.4
0.3
0.2
0.1
0
..
r~N
r.7
« May04101620.dat
May04103320.dat
^ [M ™04 1 l^sSdb^^^^^^ "
Temperature Response
A
V
^
lb~ i
I? Correct for Thermal drift?
Display Sensor Serial No default
1 '
Site Lake Harwell, OB. S.S. 5 ml/min
0 100 200 300 400
Time (sec)
500 600 700
o
Temperature (C)
ose
Peak Arrival First Moment Flow rate Flow 98;;: Confidence Limits Darcy velocity
Time fsecl Isec] [m! [ml/min! Minimum Maximum cm/dav)
Heater (2) Ch 0 |[£j£| 123.3 27.45 24.18 31.76 17.3
, fC^ 0.251 Ch4 187.5 N.A. N.A. N.A. N.A. N.A.
Intercept
75
-------
(17)
May 04 11 34 53 dat
May04114653.dat
May 04 11 5703.dat
May0412065B.dat
f^ Correct for Thermal drift?
Display Sensor Serial No default
Site Lake Hartwel, OB. S.S. 5 mVmin
Temperature Response
200 250
Time (sec)
Flow rate
Iml (ml/mini
15.20
12.413
N.A.
N.A.
Flow 98% Confidence Limits Darcy velocity
Minimum Maximum [cm/day]
13.75 17.00 3.6
N.A.
N.A.
N.A.
N.A.
NA
N.A.
(18)
* Signal Display 0[&®
Re! Temp (C deg)
ta
'._S
£
1.1-
1-
0.9
0.8
0.7
0.6-
0-5
0.4
0.3-
0.2
0.1
blien
my documents
| Flux meter
jLakeHaitwell
3^!^^^^^^^H v
/
/
May 04 11 4653.dat
May 04 11 5703.dat
May04120658.dat *
Temperature Response
,.— *~™--wfc
,/--
S
•»1
iS;S!^.
F7 Correct for Thermal drift?
Display Sensor Serial No default
1 1
Site Lake Harlraell, OB S S. 5 mlAnin
^^*""°*BS=S!a=^
0 50
0
Temperature (C)
100 150 200 250 300 350 400
Time (sec)
ose
Peak Anival Fir:,i Moment Flow rate Flow 93% Lonfidence Limits Darcy velocity
Time fsecl Isec] fm! [ml/mini Minimum Maximum (cm/dav)
Heater (2) Ch 0 [QQ 1 78.5 1 6.0S 14.52 1 S.02 1 0.1
BTeirpStQpe( -Z9E-°3 Ch 3 !££! N.A. N.A. N.A. N.A. N.A.
[C/hr]
Intercept
76
-------
(19)
* Signal Display 0@®
£3 Hen
Q) my documents
t3 Nun meter
t^LakeHartiAiell
H-^ITR^^^^^^H . .
1.1
1
0.9
0.8
SO.B
|0.5
«0.4
0.3
0.2
0.1
0
May04112253.dat
May 04 11 5703.dat
May04120658.dat v
Temperature Response
& Correct for Thermal drift?
Display Sensor Serial No default
Site Lake Hailwel, OB. S S. 5 ml/min
7
""{"
7?
/
0 50
\
a*i25;
i::===r^— ___
Q
Temperature CC)
100 150 200 250 300 350 400
Time Csec)
ose
Peak Arrival First Moment Flora ra e Flo™ 38% Confidence Limits Darcy velocity
Time fsecl Isecl fmUml/minl Minimum Maximum cmydav]
Heaterf2) ChO (JJ2JJ 213.7 11.99 10.93 13.29 7.5
[P/hrl
Intercept :
(20)
* Signal Display 0@S
Re! Temp CC deg)
•:-
,..
t
.1
1 -
19-
3.8-
3.7-
3.6'
35-
0-4
0,3-
Q2-
3.1 •
|blien
^-
"^v
— "^
0 50
f/ Correct for Thermal diift?
Display Sensor Serial No default
I I
Site Lake Harlraell, OB S S. 5 mlAnin
o
Temperature (C)
1 00 1 50 200 250 300 350 400
Time Csec)
ose
Peak Anival Fir:,i Moment Flow rate Flow 93% Lonhdence Limits Darcy velocity
Time fsecl Isec] fm! [ml/min! Minimum Maximum cm/day]
Heater (2) Ch 0 fQj£] 1 85.5 1 5.1 9 1 375 1 6.99 9.6
BTagmp'^0peef 1.1E-02 Ch3 m| N.A. N.A. N.A. N.A. N.A.
, fC^ 0.251 Ch4 S3.2 N.A. N.A. N.A. N.A. N.A.
Intercept
77
-------
(21)
* Signal Display Qfnjfx]
£3blien
Q) my documents
t3 flu* meter
H3 Lake Harwell
1.1
1 •
0.9
0.3
0.7
lo.6
o
6 °-5
Lt
0.3
0.2
0.1 -
0-
-0.1
May04112253.dat
May 04 11 34 53. dat
Temperature Response
-:
f
I
• — -
+ j.
- ^
i-
0 50 1 00
*- *u^ii
— %
"I
^-'-
I? Correct for Thermal drift?
Display Sensor Serial No default
Site Lake HaKwel, OB
^
_ :.
-^r
o
Temperature (C)
150 200 250 300 350 400 450 500 550 600
Time (sec)
ose
Peak Arrival First Moment Flow ra e Flow BQ% Confidence Limits Datcy velocity
Time (seel Isecl [ml (ml/min) Minimum Maximum cmyday]
Peak Arrival
Time fsecl
Heater [2) Ch 0 §gj|
BHHB 251 Ch1 BE3
Base Line Ref R73el
(C/hrJ 0251 Ch4 24Z3
First Moment
fsecl
N.A.
3143
N.A.
N.A.
Flora rate
[ml (ml/rnini
N.A.
4.183
N.A.
N.A.
Flora BSZ Confidence Limits
Minimum
N.A.
4.0
N.A.
N.A.
Maximum
N.A.
4371
N.A.
N.A.
Darcy velocity
ferny day]
N.A.
26
N.A.
N.A.
(22)
I 0.4
-0.1
-0.2
May0414193F.dat
May 0414 33 18 dat
May 0414 51 54.dat
MayQ4150619.dat
[v- Correct [or Thermal drift?
Display Sensor Serial No default
Site Lake Harlraell, NB
Temperature Responst
1,000
Flow 38% Confidence Limits
Minimum
N.A.
Maximum
N.A.
-6.243
N.A.
13.146
-7.335
N.A.
Darcy velocity
(cm/daul
N.A.
5.8
-4.2
N.A.
78
-------
(23)
Correct for Thermal drift?
Display | Sensor Serial No default
Site Lake HaKwel, NB, S.A. 5 ml/min
Temperature Response
(24)
May04135403.dat
M ay 04 1 41 9 35 . dat
Mw 04 1 4 51 54 dat
May04150619.dat
F/ Correct for Thermal diift?
Display SensoF Serial No default
I '
Site Lake Harlraell, NB, S A. 5 ml/min
Temperature Respons
79
-------
(25)
May 0413 54 03.dat
Mas 04 14 19 36.dat
May 04 14 3318 dal
flBmEimiaHBHi
Hay0415'lJb1Hdal
1* Correct for Thermal drift?
Display I Sensor Serial No default
Site Lake Harlwel, NB
Temperature Response
Flow 98% Confidence Limits
Maximum
2.79
N.A.
N.A.
Darcy velocity
1.6
0.3
N.A.
N.A.
(26)
May04135403.dat
May0414193E.dat
May 0414 33 18 dat
Ma»041451 54.dat
(v- Correct for Thermal drift?
Display Sensor Serial No default
Site Lake Harlraell, NB
I
0.8
0.7
i 0.6
. O.S
0.4-
0.3
Temperature Responst
Peak Arrival
Time fsecl
ChO
Chi
0.270 Ch 4 365.1
First Moment
Isecl
N.A.
221.4
170.8
Flow rate
[ml [ml/minl
N.A.
-G.715
-16.345
Flow 98% Confidence Limits
Minimum Maximum
N.A. N.A.
-6.217
-14.339
-7.296
-19.060
Darcy velocity
(cm/daul
N.A.
0.7
-4.2
-10.3
80
-------
(27)
a07
5 06
May 04 15 4447.dat
May04155442.dat
May04161324.dat
May041E3023.dat
1^ Correct for Thermal drift?
Display Sensor Serial No default
Site Lake Hartwel, NB. S.S. 5 ml/min
Temperature Response
400
450
500
550
Flora 985! Confidence Limits
Minimum Maximum
6.27 7.41
6857
N.A.
N.A.
Darcy velocity
[cm/day!
NA
N.A.
(28)
Hay 04 15 30 5E.dat
May 0415 54 42 dat
May04161324.dat
MayQ4163023.dat
(v- Correct for Thermal drift?
Display Sensor Serial No default
Site Lake Hailwel NB. S.S. 5 mlAnin
Temperature Responst
350
400
450
Peak Arrival
Time fsecl
:\ (2) Ch 0
27. S Ch1
Base Line Rel
Temp Slope "•GE-°2 ch3
First Moment
Isecl
N.A.
220.1
229.2
Flow rate
[ml [ml/minl
N.A.
-6.770
-10.998
Flow 98% Confidence Limits
Minimum
N.A.
Maximum
N.A.
282
-6.267
-3.837
242.774
-7.357
-12.511
Darcy velocity
(cm/daul
N.A.
19.0
-43
-6.3
81
-------
(29)
s Signal Display UJfelfx)
RelTemp(C deg)
~
.1
1 -
0.9-
0.8-
0.7-
0.6-
D.5-
0.4
0.3
0.2
0.1 •
0-
jblien
i my documents
;3 Hux meter
JLakeHailwell
A
f /"""i-/1
May 04 15 30 56. dal
May 04161324 dal
May04163023.dat v
Temperature Response
«rr>.>^.
^^
^^.
•-»
^N
>- -
[y Correct for Thermal drift?
Display Sensoi Serial No default
Site LakeHartwell.NE. S.S. 5mlAnin
r-.-f'^ • Y, — -=_^
0 100
200 300 400
Time (sec)
500 600 700
1
lose
PeakArrival FirstMoment Flow rate Flow 88% Confidence Limits Darcy velocity
Timelsecl (seel [ml [ml/mini Minimum Maximum cm/day]
Heater 13 Ch 0 gjjQ 2587 8.68 8.88 10.67 6.1
GTaempSfcjpe8' 9-5E-°2 Ctl 3 E£] NA NA NA NA NA
[C/hrl 02/7 CM NA NA NA NA NA
Intercept
(30)
* Signal Display MIMS
RelTempCC deg)
(z,
~
f
i
•
.1-
0.9-
0.8-
0.7-
0.6-
0.5-
0.4
0.3-
0.2-
0.1-
0
jblien
3 my documents
^ [lux meter
^jLakeHailwell
_j|^3^H^^H v-
/
_ - /_ —
May 041530 56. dal
May04154447.dat
May 04 16 50 23 dal »
Temperature Response
0 100
5^1
200 3
PeakArrival First Mo
Time [seel fsec
Heater (2) Ch 0 J^jJJ 373.
. lCyhr! 0.273 Ch 4 60.3 N.A
Intercept
~^^
17 Correct [or Thermal drill?
Display Sensoi Serial No delault
Site Lake Hartwel. MB. S.S. 5 ml/min
•+
oi
I
ose
0 400 500 600 700
Time (sec)
lent Flow rate Flow 88% Confidence Limits Darcy velocity
[ml (ml/mini Minimum Maximum cm/dayl
5 6.15 5.68 6.70 3.8
3 4.588 4.4 5.523 2.8
N.A. N.A. N.A. N.A.
N.A. NA N.A. N.A.
82
-------
(31)
s Signal Display UJfelfx)
R el Temp (C deg)
~
.1
1
0.9-
0.7
0.6'
0.5-
0.4-
0.3-
0.2
0.1 •
0'
jblien
imy documents
^ Nun metet
iJLakeHartaell
_J^^^^H ^| v
/
•Y
: J*
Maj 04 15 30 56. dot
May04154447.dat
May 041554 42. dat
Temperature Response
.^rrT"4_
^
0 100
-
s,
li* Correct for Thermal drift?
Display Sensoi Serial No default
Site LakeHartwell.NE. S.S. 5mlAnin
»**.
200 300 400
Time (sec)
500 600 700
0
Temperature (C)
lose
PeakAiriva! FirstMoment Flow rate Flow 88% Confidence Limits Darcy velocity
Time [sec] [sec] [ml [ml/mini Minimum Maximum cm/daii]
Heater(2) Ch 0 gjjjjj 377.8 6.17 5.71 6.72 3.9
^^^^^| 27.3 Ch1 f{£]J 286.2 4.722 4.5 5.707 3.0
GTempSfcjpe8' -1-9E-u2 Ch 3 (jjjjij 78-8 -43.000 -36.658 -51.876 -27.0
[C/hr] 02?g C(l4 64g NA NA NA NA NA
Intercept
(32)
* Signal Display MIMS
RelTempfC deg)
^
~
f
i
•
i.i-
1
0.9-
0.8-
0.7-
0.6
0.5-
0.4
0.3-
0.2-
0.1 •
0-
Men
3 my documents
^ Nun meter
^jLakeHailmell
_j|^3^H^^H v
L-
jt
yr
/
1
May04171542.dal I I Close
Maj 04 17 23 23 dat DBpteji Sensoj Serial No default
Maj04174022.dat ' '
Maj04174842.dal v Site Lake Hartwel. NB. S.S.5mlAnin
Temperature Response
^^
r*~-
- - ^*—
0 100
-'""SS^
~,
200 3
PeakAttival First Mo
Timefsec) fsec
Heater (2) Ch 0 |fg£| 389.
MmX6' 2-8E-°2C" la
. lCyhrI 0.273 Ch4 190.4 N.A
Intercept
—-- -
-
0
Temperature (C)
0 400 500 600 700
Time (sec)
lent Row rate Flow 88^ Confidence Limits Darcy velocity
[ml (ml/mini Minimum Maximum cm/daii)
5.87 5.53 6.50 3.8
1 4.560 4.3 5.484 2.8
N.A. N.A. N.A. N.A.
NA NA N.A. N.A.
83
-------
Pj my documents
meter
Maj04172929.dat
May 041740 22. dat
May04174842.dal
ff Correct for Thermal drift?
Display Sensor Serial No default
Site Lake Hartwell. NB S.A. 5 ml/min
Temperature Response
(34)
* Signal Display 0@S
el Temp (C deg)
^
a
•
1.1 •
1 •
0.9
0.8
0.7
0.6-
0.5-
0.4
0.3
0.2-
0.1
0-
0.1
•0.2
0.3
blien
my documents
j5G£^H^^^^I v
•
May 04 16 47 22. dat
May 041740 22. dal
May 04 17 48 42 dal »
Temperature Response
;--/^L>.
i— -
0 100
i ii- '^"
P Correct for Thermal drift?
Display Sensor Serial No default L
Site Lake Hartwel, NB. SA. 5 ml/min
. •
A ;
"*'™***il*"'!"iii 'v \il; -
*••• *St.j*»' li
t«:« ;IJL
"::"rr:";Z^|^
200 300 400
Time (sec)
0
Temperature (C)
500 600 700
ose
Peak Arrival Fist Moment Flora rate Flora 98* Confidence Limits Darcy velocity
Time [sec] [sec] [ml ImlAnin] Minimum Maximum cm/dav]
Heater (2] Ch 0 [fjff N.A. NA N.A. NA. N.A.
^SStmf 86E-02 Ch 9 mSSOSm N.A. N.A. NA. N.A. N.A.
84
-------
(35)
s Signal Display | __|| ^|fx)
Rel Temp (C deg)
_1
6
~
1.1 •
i-
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
-0.1
-0.2.
blien
my documents
3 flux meter
SLakeHarlwell
JQE^^^H
..-
J>
r *
tf»
MayQ417154Zdat
^— M*ftpQ4174B42.dal
Temperature Response
jf £•
•?*-*•.;-
v3*
& Correct for Thermal drift?
Display Sensor Serial No default
Site Lake Hartwel. NB S.A. 5mlAnin
|
0 50 100 150 200
Time (sec)
1,
•';
^L •*-
^*
*~-=.
250 300 350
0
Temperature CC)
lose
PeakArrival FirstMoment Flow rate Flow 38% Confidence Limits Darcy velocity
Timelsecl (seel [ml [ml/mini Minimum Maximum cm/daiil
Heater (2) Ch 0 f^Q N.A. N.A. N.A. N A. N.A.
GTaempSfcjpe8' Z4E-°3 Ctl 3 EM NA NA NA NA NA
[C/hrl
Intercept
t^ blien
^ my documents
^3 Nun meter
llakeHartwell
May 04 18 05 42 dat
May04181258.dat
Maj.04183001.dal
* Correct for Thermal drift?
Sensoi Serial No default
Site Lake Hartwel. NB. S.A. 5 ml/min
Temperature Response
BTemp™opee'
I0*'1 0.238
Intercept
-1.630
-4.012
-1.513
-3.723
-1.756
-4.371
-1.0
-2.5
85
-------
Pj my documents
<^ flux meter
a Lake Harwell
May 041740 22.dal
Hay04174S42.dal
Mav 04181258 dot
May 0418 30 01.dal
ff Collect for Thermal drift?
Display Sensor Serial No default
Site Lake Hartwell. NB S.A. 5 ml/min
Temperature Response
First Moment Flow rate Flow 38% Confidence Limits
(sec] [ml [ml/mini Minimum Maximum
N.A. N.A. N.A. NA
225.4
-1820.7
-6.553
-0.121
-6.075
-0.197
-7.123
-0.030
Darcy velocity
Icm/daul
N.A.
N.A.
-4.1
-01
(38)
* Signal Display
pay04174022.dat
May04174842.dat
Hay 041805 42.dal
IjBlliEibihi&BM
May 04 18 30 01.rial
p Correct for Thermal drift?
Display Sensor Serial No default
Site Lake Hartwell, NB
Temperature Response
400
Peak Arrival
Time [seel
Heater 12}
ChO
Base Line Fief O.cn, r. ,
Temp Slope -&4E-03 Ch3
[C/hrl 0 2M Ch 4
Intercept
First Moment
Isecl
327.6
270.3
31.6
N.A.
Flora rate
1ml [mlAnin]
7.27
5.076
N.A.
N.A.
Flow*': Lurrlidence Limits
Minimum
6.70
NA.
NA.
Maximum
7.35
N.A.
N.A.
Darcy velocity
fcm/dav]
4.6
3.2
N.A.
N.A.
86
-------
(39)
s Signal Display UJ 1° Ifx)
R el Temp (C deg)
J
6
~
1.1 •
1 •
0.9
0.8
0.7
OS-
0.5
0.4
0.3
0.2
0.1
0-
0.1
-0.2
mv documents
3 Nun meter
jjLakeHartwell
J^^^^H ^| v
"f
"
May04174022.dal
May 04 17 48 42. dat
May 04 18 05 42. dat
Temperature Response
J*
X
0 50 100
*+~*
P«V*
**r
*•«».
"^»«^
^
^
'""jiii<«i
yA
K> Correct for Thermal drift?
Display Sensoi Serial No delault
Site Lake Hartwell, ME. SA. 10 rnl/mtn
.#*;_
*
£s
-p*
j^fr
»5sr
K"'.
0
Temperature (C)
150 200 250 300 350 400 450 500 550 600
Time (sec)
lose
PeakArrival FrrstMoment Flow rate Flow 98% Confidence Limits Darcy velocity
Time [sec] [sec] [ml [ml/mini Minimum Maamum cm/daii]
Heater (2) Ch 0 |£j£jj N.A. N.A. N.A. NA. N.A.
^empSfcjpe8' Z4E'02 c" 3 |
------- |