Hft.
EPA-600/4-81-005
February 1981
DEVELOPMENT OF A HELICOPTER
WATER QUALITY MONITORING/SAMPLING SYSTEM
by
H. Michael Lowry
Advanced Monitoring Systems Division
Environmental Monitoring Systems Laboratory
Las Vegas, Nevada 89114
EJBD
ARCHIVE
EPA
600-
4-
81- ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
005 OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89114
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring Systems
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
-------
Repository Material
Permanent Collection
SUMMARY
USEPA-EMSL-LV has developed a helicopter borne water quality monitoring/
pump-sampling system for sampling lakes and rivers. Development of this
sampling system originated at the beginning of the National Eutrophication
Survey (NES) in May of 1972 for use in Bell UH-1H, "Huey" helicopters
(outfitted with floats), but the design concepts apply to any aircraft.
Development continued intermittently, while the system was in use, for the
next 5 years. This document reports the problems encountered in our original
sampling effort and the design concepts applied to correct those problems.
The basic system, originally designed for ship-board use (InterOcean
model 513 with pump sampler), was mechanically tailored to fit our aircraft.
Components were replaced or modified as deficiencies became apparent.
Water samples of any volume can be collected to a maximum depth of 172
meters at a rate of 7 liters/minute and in-situ measurements can be made of
conductivity, temperature, sampling/measuring depth, dissolved oxygen, pH, and
turbidity.
From our experiences with this sampling system we have concluded that:
-- helicopters equipped with floats are versatile, cost-effective
sampling platforms;
-- this sampling/measuring system is reliable when properly maintained;
-- the pump-sampling concept is the desirable way to collect multiple
samples from multiple depths in the water column to a depth of about
350 meters;
-- electrical interferences and vibration problems are hazards when
5? using water quality monitoring systems in an aircraft;
to
-- dissolved oxygen and pH sensors are extremely vulnerable to mechanical
damage and electrical interference, and are of questionable utility
for use with an airborne system; and
~ to ensure continued reliable operation, water quality monitoring
systems require frequent calibration and maintenance.
u * USEPA
Headquarters and Chemical Libraries
EPA West Bldg Room 3340
711 Mailcode 3404T
1301 Constitution Ave NW
Washington DC 20004
202-566-0556
-------
iv
-------
CONTENTS
Summary Ill
Figures vii
Tables viii
Abbreviations and Symbols ix
Introduction 1
Conclusions 2
Recommendations 3
General recommendations 3
Specific recommendations 3
Water Quality Sampling Platforms 5
Sampling platform requirements and options 5
Helicopter sampling platform 6
Helicopter-Borne Water Sampling System 9
In-situ sensors and pump 9
Assembly 9
Junction box 12
Pump 12
Sensor and analog electronics 15
Hose/cable, boom, sheave and winch 20
Hose/Cable 20
Boom 21
Sheave 21
Winch 21
Onboard electronics 25
Mounting rack 25
Console 25
Graphical displays 25
Winch controller 27
Pump controller 27
Power supplies 27
Echo sounder 27
Sampling manifold and sample storage 28
Other underwater components 28
Solar illuminance meter and secchi disk 28
Sediment sampler 29
Echo sounder transducer 29
-------
CONTENTS (Continued)
Page
Sampling Operations 30
Calibration and Maintenance 32
Sampling Capabilities 36
Discussion 38
Appendices
A. The Dissolved Oxygen Circuit 41
B. The Differential-Input pH-Measuring Circuit 44
C. Water Column for Calibration of Echo Sounder 47
D. Calibration Buffer Board 49
Literature Cited 51
VI
-------
FIGURES
Number Page
1 In-situ sensor/pump assembly being deployed from the "Huey"
helicopter while on station 7
2 Floor plan of the helicopter with the monitoring/sampling
system located 8
3 Block diagram of the helicopter water quality monitoring/
sampling system 10
4 Plan view of the in-situ sensor/pump assembly showing
relocation of the turbidity sensor 11
5 In-situ sensor/pump assembly with components identified ..... 13
6 Hypro model B-612 positive displacement pump 14
7 Inductive conductivity sensor (InterOcean) 16
8 Dissolved oxygen sensor (Yellow Springs Instruments) 18
9 Process Lazaran reference and glass pH-measuring electrodes
(Beckman) 19
10 Hinged boom mounted in the helicopter 22
11 Light-weight sheave for the sensor hose/cable 23
12 New 100-meter winch mounted in the helicopter 24
13 Onboard electronics subsystems mounted in the helicopter 26
14 Dye study of helicopter rotorwash-turbulence effects on
Lake Mead 40
vii
-------
TABLES
Number page
1 The Parameters and Types of Sensors Used on the InterOcean
In-Situ Sensor System 9
2 Daily Calibration Procedure for Helicopter Borne Water
Quality Monitoring Package 33
3 Daily Operational Checks of the In-Situ Sensor/Sampling
Package 35
4 Specifications of the In-Situ Water Quality Monitoring
Package 37
vm
-------
ABBREVIATIONS AND SYMBOLS
A — ampere
a.c. -- alternating current
cm — centimeter
d.c. -- direct current
DO -- dissolved oxygen
ft -- feet
FTU -- formazine turbidity unit
g -- gram
gpm -- gallons per minute
gal -- gallon
hp — horsepov/er
in -- inch
10 -- InterOcean
kg -- kilogram
kw — kilowatt
1 — liter
1pm -- liters per minute
Ib -- pound
m — meter
mm — millimeter
psi ~ pounds/square inch
rpm -- revolutions per minute
V -- volt
VA -- volt ampere
-------
INTRODUCTION
In the spring of 1972, the U.S. Environmental Protection Agency's
Environmental Monitoring Systems Laboratory at Las Vegas (EMSL-LV) initiated
the National Eutrophication Survey (NES), a 4-year field investigation of the
effects of nutrients on the aging process of the nation's lakes and
reservoirs. The program was designed to sample each of 815 lakes, distributed
nationwide, once each during spring, summer, and fall seasons (USEPA 1974,
1975). It was necessary to collect water samples for laboratory analyses that
would best characterize the individual water bodies in terms of nutrient
levels and their manifestations, e.g., chlorophyll ^concentrations and
phytoplankton composition.
To sample this large geographical area in 4 years and to visit each lake
three times in the same year, aircraft were considered the only practical
survey platform. After some field experience with both helicopters and
fixed-wing aircraft, helicopters were chosen as the most versatile and
cost-effective survey platform to accomplish our mission. No commercial water
quality monitoring/sampling system designed specifically for use from an
aircraft was available, nor was there time for system development before
beginning the field effort. To execute tha rigorous schedule of the field
study, a ship-board system was procured and subsequently modified for airborne
application during the survey.
This report describes the helicopter water quality sampling system
developed originally for use by the NES and subsequently used in support of
other water quality projects. It also describes the salient problems
associated with the use of a helicopter as a sampling platform and the
modifications made in the system to fulfill our needs. This system should be
useful to other groups involved in sampling large numbers of water quality
sites distributed over large or inaccessible areas.
-------
CONCLUSIONS
1. Our experiences indicate that helicopters equipped with floats are
versatile sampling platforms and are cost-effective when large or
inaccessible geographic areas must be sampled quickly with few personnel.
2. The sampling/measuring system described here is reliable when properly
maintained.
3. The pump-sampling system is desirable when multiple water samples of
variable volume are to be collected from many different depths in the
water column (to a maximum of about 350 meters), especially from sampling
platforms with limited space.
4. Electrical interference and vibration problems are hazards when using
water quality monitoring systems in aircraft.
5. Outfitting an aircraft for water sampling requires detailed planning and
engineering.
6. Dissolved Oxygen and pH sensors are extremely fragile, easily fouled, and
vulnerable to electrical interference and "ground-loops." These sensors
and circuitry require constant attention by qualified technical personnel
to ensure reliable dissolved oxygen and pH data.
7. To ensure continued, reliable operation in the field, water quality
monitoring systems require frequent maintenance and calibration.
-------
RECOMMENDATIONS
GENERAL RECOMMENDATIONS
1. Helicopters are recommended for use as water sampling platforms when
large numbers of lakes or lakes located in inaccessible areas must be
sampled quickly with few personnel.
2. This helicopter-borne water monitoring system is recommended when there
is sufficient space in the aircraft and sufficient electrical power
available.
3. When electronic water quality monitoring systems are used, a well
experienced electronics technician with a high mechanical aptitude and a
general chemistry background should be part of the field team.
SPECIFIC RECOMMENDATIONS
These recommendations may not be appropriate for all systems; they are
based on our experiences with our system. They were implemented and tested on
our system or will be implemented on the next generation system.
1. Because of their maneuverability on the water and ability to take-off and
land in a limited area, helicopters are recommended over fixed-wing
aircraft as water sampling platforms.
2. Static inverters are recommended where 117 V a.c. power is required
(the power factor should be corrected to one).
3. Aircraft power should be used vnthout conversion wherever possible.
4. Equipment racks should be designed to conform to the internal contours of
the craft.
5. Sample storage racks should be designed, where possible, to utilize space
that is otherwise unusable.
6. All equipment and storage racks should be secured to the aircraft.
7. All permanently mounted hardware should be shock-mounted. Shock mounts
should conform to military specification Mil. Std. - 810. All deploy-
able systems should have cushioned restraining cradles for on-board
storage. All fragile electrodes should be individually shock mounted.
-------
8. Hardware for in-situ measurements and sampling should be as light-weight
and compact as possible.
9. Strong, light-weight, noncorrosive materials (i.e., plastics) should be
used wherever possible, especially on deployable hardware.
-------
WATER QUALITY SAMPLING PLATFORMS
SAMPLING PLATFORM REQUIREMENTS AMD OPTIONS
The NES sampling program required visits to several hundred lakes
scattered across the 48 conterminous United States. Even with optimal
planning of the sampling schedule, distances as great as 160 km (100 miles)
separated consecutive lakes. Each lake was to be sampled a minimum of three
times and, generally, at more than one sampling site. In-situ measurements
and water samples were to be collected from various depths in the water column
for physical, chemical and biological analyses. The magnitude of the program
required that the sampling platform be a) large enough to safely ferry the
sampling crew and the sampling system; b) stable and easy to work from on the
water; c) capable of providing sufficient electrical power to operate the
sampling system; d) capable of maintaining position at the site during
sampling operations in wind and water currents; e) capable of sampling an
average of eight sites per day in a 100-mile radius; f) reliable; and g) cost-
effective.
The options for sampling platforms considered included small boats,
fixed-wing aircraft, and helicopters. To meet the time constraints, use of
small boats would require a large number of independent field crews and
monitoring/sampling systems. This would be an ineffective use of manpower and
too expensive. Aircraft represented a much more attractive option.
The selection of the type of aircraft was based on several considera-
tions. Although fixed-wing craft are more economical to operate and maintain
than helicopters, and have a greater operational radius in terms of time and
distance, they require more open water for take-off and landing, and their
maneuverability on the water is minimal. Also, the doors on fixed-wing craft
are too small for convenient deployment and recovery of sampling packages and
the standard floats used are not configured to accommodate personnel during
sampling operations. Helicopters, although less economical to operate and
more limited in operational radius, can take-off and land with minimal open
area and are capable of lateral as well as fore and aft maneuvers, and
rotation about a central vertical axis while on the water. Floats are
available that have a flat, deck-like top surface for working personnel. Some
aircraft have large sliding doors convenient for the ingress and egress of
sampling equipment. The more reliable helicopters are powered by jet
turbines. However, the number of small airports that can supply jet fuel,
particularly among those in rural or remote areas, are limited. If a
turbine-powered craft is selected, fuel may have to be trucked into the
sampling area, or some intermediate location to minimize the time lost in
refueling.
-------
Bell UH-1H "Huey" helicopters and one de Havilland U-la "Otter"
fixed-wing aircraft were borrowed from the U.S. Department of Army. All three
craft were outfitted with floats and used in the field during the first month
of the NES. Because it required too much open water for take-offs and could
not maneuver on the water to maintain station, the "Otter" was used only for
cargo and personnel transport. The "Huey" was quite maneuverable and
versatile and, after the first month of sampling, was chosen to complete the
survey. A third "Huey" was obtained from the Army.
HELICOPTER SAMPLING PLATFORM
The three Bell UH-1H "Huey" helicopters (Figure 1) used as sampling
platforms are thirteen-passenger craft with a total payload of 1,400 kg (3,100
Ibs) at sea level, and are powered by a single jet turbine delivering 320-kw
(mechanical) (1,100-shaft horsepower). The total available electrical power
to operate sampling equipment is approximately 200 A at 28 V d.c. or 5,600
watts. Each aircraft was fitted with floats (a standard Bell retrofit kit) to
allow it to land on the lake surface (as opposed to hovering) thereby
conserving fuel, reducing the risk of injury, and providing greater sampling
convenience.
The helicopter's main fuel tank capacity is 635 kg (1,400 Ibs), yielding
about 2 hours of flying time at 100 knots (the airspeed is limited to 100
knots by the floats). After the first year of the NES, an auxiliary fuel tank
with a capacity of 136 kg (300 Ibs) was installed in each aircraft. This
increased the flying time by 1/2 hour. While performing sampling on the water
surface, the turbine was idled to reduce fuel consumption (the rate at flight
idle is about 1/2 the flying consumption rate) and yet retain sufficient power
to maintain position on the station. The total time between refueling stops
was in excess of 2-1/2 hours since each half hour of sampling consumed the
equivalent of only 15 flying-minutes of fuel.
All of the seats except those of the pilot and co-pilot were removed to
accommodate the auxiliary fuel tank, sampling equipment and sample storage.
The auxiliary tank served as a bench seat for the sampling technician and any
additional passengers. The arrangement of the sampling equipment and storage
is shown in the floor plan of the cabin (Figure 2).
-------
lsggt-BCT'WPfMSBi;
:^^^
^.At^ •'^•ffn-fH-vt ™^M%i,, " J3»-' .T
1
/Figure 1 In-Xitux^ensor^ump Assembly Being Deployed from the "Huey" ^felicopter ^hile^On Ration.
^
-------
Figure 2. Floor plan of the helicopter with the monitoring/sampling system located,
-------
HELICOPTER-BORNE WATER SAMPLING SYSTEM
Each of the three Hueys was outfitted with a sampling system. The
sampling system consists of six in-situ sensors (Table 1), their associated
analog electronics, and a submerged pump, mounted together on a frame; a hose/
cable, boom assembly and winch; on-board electronics; a solar illuminance
meter, a Secchi disk and a sediment sampler (Figure 3).
TABLE 1. THE PARAMETERS AMD TYPES OF SENSORS USED ON THE INTEROCEAN
IN-SITU SENSOR SYSTEM
Parameter Type of Sensor
Conductivity Inductive sensor
Temperature Glass-bead thermistor
Depth Wheatstone bridge
strain gauge
pressure transducer
Dissolved Oxygen Polarographic membrane
w/external temperature
compensation
pH Mon-ruggedized combination
pH electrode w/external temperature
compensation
Turbidity Optical transmissometer
IN-SITU SENSORS AND PUMP
Assembly
The in-situ sensor/pump sampling system was of a commercial ship-board
design for use in a marine environment and manufactured by InterOcean of San
Diego, California (Model 500, modified to include a pump sampling system). It
-------
POWER lllllilltiE»9191
WATER
SIGNAL
Figure 3. Block diagram of the helicopter water quality
monitoring/sampling system.
10
-------
was heavier and bulkier than necessary for the acquisition of data from most
inland water bodies. Both mass and bulk are critical in airborne applica-
tions. The in-situ sensor assembly consists of five sensors mounted on the
end of a cylindrical stainless steel pressure case (10 cm diameter, 41.3 cm
length, 8.5 kg weight) and a sixth optical sensor mounted on the side of the
pressure case. The case housed the analog electronics and calibration
controls for the sensors. The pressure case and pump were mounted on a
stainless steel frame 30 cm (12 in) in diameter and 96 cm (38 in) in length.
The frame was mechanically attached to the hose/cable via a Kellem"" grip. The
signal and power cables, and the hose "fanned-out" from a potted splice on the
end of the hose/cable.
Handling of the sensor package during sampling operations was initially
the source of some sensor damage. In particular, the dissolved oxygen sensor
and the pH sensor were occasionally broken and the alignment of the
transmissometer optics was disturbed. Reduction of the size and weight of the
probe-assembly and relocation of the turbidity sensor in a more protected
location, closer to the center of the assembly, reduced the vulnerability of
the sensor to rough handling (Figure 4).
Sensor
Pressure
Case
Original \
Location
Turbidity J
Final
Location
Turbidity
Pump
&
Motor
Figure 4. Plan view of the in-situ sensor/pump assembly showing relocation
of the turbidity sensor.
11
-------
Junction Box
Originally, there was no provision for strain relief where the pigtails
to individual components on the sensor/pump assembly exited the potted splice.
The constant motion of the wires at the splice caused some conductors to fail.
Repair was not readily effected since the potting compound had to be cut away,
the pigtails respliced to very short wires at the pot, and the splice resealed
to maintain watertight integrity. InterOcean provided the solution to this
problem. The multiconductor cable/hose was brought into a watertight oil-
filled junction box. The box was fitted with rubber-bushed plastic stuffing-
glands for cable entry and a rubber diaphragm in one side for pressure
equalization. All splices are now inside the junction box and are easily
repaired or modified. The box is mounted firmly to the frame so the cables
can no longer move and be broken (Figure 5).
Pump
The pump originally supplied by InterOcean was a 250-watt (mechanical)
(1/3 horsepower) 11-stage centrifugal well pump of cast naval bronze. It
weighed 16 kg (35 Ibs) and was 80 cm (32 in) in length and 9.5 cm (3 3/4 in)
in diameter. The motor was manufactured by Franklin Electric Company. The
pump delivered about 7.5 1pm (2 gpm) through a 100-meter length of hose/
cable. The delivery rate is inversely proportional to the total length of
hose; it is not affected by sampling depth. There was a protective screen
over the intake that would pass nothing larger than approximately 1.6 mm
(1/16 in) in diameter. The screen blocked the passage of algal masses and
jeopardized quantitative estimates of chlorophyll a_ and phytoplankton
components. When the screen was removed, accumulated algae eventually clogged
the pump impellers and stopped the pump. The length of the pump was the
limiting factor in reducing the size of the sensor/pump assembly.
Since a smaller, lighter, positive-displacement pump would allow
reduction in the size and weight of the assembly and avoid the clogging
problems, such a pump, the Hypro Flexroller model B-612, which weighs only 2
pounds without the motor, was procured. It required very low start-up torque,
and was designed for a maximum operating speed of 1,800 rpm. The pumping seal
was maintained by four rubber rollers (Figure 6). The only sealed submersible
motor available that met size, mechanical, and electrical power requirements
was the motor from the original pump, which ran at 3,400 rpm. The rubber
rollers in the pump were replaced with Teflon rollers, and the pump and motor
combination was tested with the hose/cable.
The new pump, operating at 3,400 rpm, delivered more than 34 1pm (9 gpm)
at zero meters of head. However, when 34 1pm of water was forced through 100
m (328 ft) of 9.52-mm (3/8-in) tubing, the friction head loss in the tubing
was greater than 300 m (1,000 ft). In other terms, the pressure differential
from one end of the hose to the other was about 30 kg/cm* (430 psi). This was
more pressure than the hose could withstand and required more power than a
250-watt (1/3-hp) motor could deliver. The solution was to relieve the
pressure and reduce the flow accordingly. A "bleed-hole" was drilled in the
12
-------
Sheave
Junction
Pressure Case r
\ .1- - "• ••»•••'
\ ' ~ - «; -i.«i
. Flexible Coupling
Figure 5 In^itu Censor/Pump Assembly with Components..Identified.
-------
BKTH
-A--.
Figure 6. Hypro model B-612 positive displacement pump shown on the'left 1s Identical intern-ally to the
model B-634 shown in cut-away on the right,
-------
output of the pump and was enlarged until the pressure at the pump was reduced
to 7 kg/cm2 (100 psi). This diverted 3/4 of the pumped water back to the lake
and yielded a flow rate of about 7.5 liters (2 gal) per minute which was
acceptable for sampling. A better solution might have been a slower motor
(1,800 rpm) or a gear reducer between the motor and the pump. Neither was
readily available in a compact watertight assembly and the method used
represented a simple, cost-effective solution to the problem.
After modification the smaller, lighter pump assembly (Figure 5) weighed
8.6 kg (19 Ibs) and v/as 46 cm (18 in) long vs. the 16 kg (35 Ibs) and 31 cm
(32 in) of the original assembly (both weights include motor). Thus a new,
smaller frame assembly, 30 cm (12 in) in diameter by 81 cm (32 in) in length,
was fabricated from aluminum (Figure 5), which provided a significant
reduction in weight and bulk. The frame could have been made smaller but
these dimensions provided some sensor protection and facilitated handling
during deployment and retrieval.
Sensor and Analog Electronics
Conduct!vity--
The conductivity sensor used is an inductive type, designed and
fabricated by InterOcean that consists of two torroidal coils molded together
in a donut shape (Figure 7). The pair of coils acts as a transformer. The
electromagnetic coupling between the coils is a function of the conductivity
of the water column that passes through the "donut." One coil is driven with
a "saw tooth signal" generated by an oscillator; the amplitude of the signal
in the second (sensor) coil is processed to yield a d.c. voltage analog of
conductivity. This type of sensor is not susceptible to electrode
contamination. Only one problem was experienced with this conductivity-
measuring device; the driver oscillator frequently would not initiate
oscillation at temperatures below 0°C. The value of the resistor that
provided the forward bias to the oscillator transistors was changed to move
the operating region to a more reliable range on the "characteristic curve" of
the transistors.
Temperature--
The response time of the InterOcean glass bead-thermistor temperature
sensor was too long to allow rapid profiling of the water column. The
thermistor and its electronics were replaced by a platinum resistance
thermometer (originally manufactured by Rosemount and adapted by InterOcean)
and appropriate analog electronics (InterOcean retrofit). The time constant
for the thermometer is 60 milliseconds, more than a factor of 50 faster than
the thermistor, and thereby allowed much more rapid temperature profiling.
The platinum resistance thermometer sensor cost $485.00 each vs. $4.00 for the
thermistor at the time of purchase in 1972.
Depth--
The depth sensor is a bonded strain-gauge pressure transducer; the
measuring element is a resistive Wheatstone bridge. Some units were
temperature sensitive, and changed calibration with a temperature change.
Some sensors exhibited as much as 1/2 foot per degree Celsius drift. Those
sensors with extreme drift were replaced by the manufacturer. As a backup
15
-------
iaure 7. :' &
^^
,v.
s
-------
means of obtaining measurements of the depth of sampling, the hose/cable was
marked at 5-foot intervals with color-coded tape. Visual interpolation was
employed when the depth value was located between the markings.
Dissolved Oxygen (DO)--
The response times of the dissolved oxygen sensor (Beckman 39552) and
associated temperature-compensation thermistor were too long (in excess of 5
seconds) to allow rapid profiling. The change in the sensor and circuitry for
temperature allowed and improved dissolved oxygen sensor response time. By
utilizing the analog output voltage of the temperature circuitry rather than
the thermistors to provide temperature compensation information for the oxygen
circuitry, the response time of the dissolved oxygen information was reduced
to about 4 seconds (Appendix A). The oxygen sensor provided by InterOcean had
two advantages over other sensor designs: 1) the low volume and low thermal
mass yielded quicker responses to temperature and oxygen changes than other
oxygen sensors, and 2) the low oxygen-consumption rate eliminated the need for
a stirring mechanism. However, the sensor was very fragile and developed
internal electrolyte leaks that yielded inflated oxygen readings. Vie were
forced to change to a more rugged sensor, manufactured by Yellow Springs
Instrument Company, which had a slower thermal response. The oxygen sensor
provided by Yellow Springs (YSI) is model YSI 5419, a pressure compensated
sensor that YSI installed in a stainless steel fitting provided by EPA (Figure
8). The sensor has a body of PVC and acrylic, and silver and gold electrodes;
it is a clark-type polarographic electrode. InterOcean has since gone to yet
another supplier for DO sensors.
There are still inherent problems with the new DO sensor. The sensor
membrane is easily torn and subject to fouling by algae and sediment. The DO
circuitry is vulnerable to electrical interference and "ground-!oops;" the
polarizing potential applied to the sensor can, at the same time, interfere
with the pH signal. Both sensor and circuitry require frequent attention by
the technician. For these reasons, we found this sensor to have limited
utility in the helicopter application.
pH-
The combination electrode for pH was the most fragile of all the sensors.
InterOcean redesigned the sensor mounting to improve its resistance to shock,
but the electrode still failed mechanically in our application. Failures were
attributable to both extreme shock (such as dropping the sensor package) and
to aircraft vibration.
The circuitry provided by InterOcean for the pH analog was quite
vulnerable to electrical interferences and ground loops from other sensors and
the aircraft. We redesigned this circuit utilizing a differential amplifier
(Appendix B), and separate Beckman Lazaran reference and glass measuring
process-type electrodes (Figure 9) that resulted in a much more rugged sensor
assembly and more stable, interference-free electronics. InterOcean now uses
a similar differential amplifier circuit to reduce interference.
The modified circuit of the extremely high-input-impedence amplifier
required by pH sensors (>1012 ohms) is less vulnerable to "ground-loops" and
electrical interference. Very clean water (specific conductance <200
17
-------
Stainless Steel Fitting (Screws into Pressure Case)
Pressure Equilizer
\
'O' Ring
Membrane
Retainer
Membrane
Protective Cap
Figure,8 Dissolved
How Springs Instruments).
-------
Stainless Steel Fitting (Screws into Pressure Case)
Beckman
Stainless Steel Electrode Housing
Containing Buffer Preamp
(La za ra n|®]jf eck ma nj
iv-n Q
r*r^o«G3x-^*!
/~
-------
ymhos/cm) does not ionize sufficiently for the p'H sensor to be effective.
When the sensor package is not in the water, a protective cover filled with
water should be placed over the glass measuring electrode to keep the
electrode v/et. If dry, the sensor may have to be soaked for 2 to 3 hours
before consistent pH values can be obtained on calibration.
Turbidity--
The turbidity sensor used, is a simple transmissometer consisting of a
light bulb, a photo cell, and associated optical system. The output signal
is an analog of the percentage of light transmitted along a fixed 10-cni path
length through the water. This is displayed on the digital display. The
analog meter is labeled in formazine turbidity units (FTU), however, the FTU
values do not correlate with Jackson Turbidity Units (JTU) in all cases. The
only problems with this sensor were those of mechanical misalignment of the
optics caused by rough handling and its relative lack of precision.
HOSE/CABLE, BOOM, SHEAVE AND WINCH
The in-situ package mechanical delivery/recovery system that consisted
of a hose/cable, boom, sheave and winch from InterOcean provided the means to
recover water samples from discrete depths, supply electrical power for the
sensors and pump, and transmit the analog signals (Figure 3).
Hose/Cable
The hose/cable, a concentric structure, was constructed with a 10 mm
(3/8 in) i.d. nylon sample delivery hose at the center. Twenty conductors of
20 gauge stranded, Teflon insulated wire were wound around the hose at about
one turn per 60 cm (2 ft), and the voids between the conductors were filled
with a gelatinous substance to minimize water movement in this layer if the
cable inner jacket were damaged. These power and signal transfer conductors
were covered with a single, concentric vinyl jacket, and a galvanized steel
wire strain member was braided over the jacket to support the mechanical load
of the in-situ sensor/sampler system. The voids in the braid were not filled,
and the braid was covered with a second vinyl outer jacket about 0.305 mm
(0.012 in) in thickness.
In use, this thin vinyl outer jacket was easily damaged and the braid
exposed. Water from the outside then moved along the strain-member layer.
The natural working of the braid strands against each other wore away the
galvanized coating, which allowed the steel braid to rust and ultimately led
to strain-member failure. Fortunately, the conductors and the hose were
capable of supporting the relatively light load of the sensor package. When
the cable-design oversight was detected, we specified and procured new cable
with a stainless steel braid instead of the original galvanized steel and a
much thicker (2.4-mm) outer jacket. We also specified a smaller bend radius
(20.3 cm (8 in) vs. 35.5 cm (14 in)) than that of the original cable, in order
to reduce the diameter of the sheave on the boom to 40.6 cm (16 in). The
original hose/cables were each 100 meters (328 ft) long. The new cables were
100 and 200 meters long to fill the new winches.
20
-------
Boom
The original boom was mounted on a pivoting mast that attached to
existing fittings in the deck and the overhead in the craft. After the first
year, a new hinged-boom system was designed which attached to three existing
attachment points in the "overhead" and used a fourth point for bracing in the
extended mode (Figure 10). This configuration allowed relocation of the
winch a few centimeters forward and outboard of its original position, which
in turn, made available the necessary room to add an auxiliary fuel tank.
Sheave
The sheave provided on the original procurement was a massive aluminum
casting more appropriate for marine applications. For our relatively short
cable and cramped quarters onboard the aircraft, a smaller, lighter sheave was
needed. This need was satisfied by casting polyurethane in the modified rim
of a heavily constructed 40.6-cm (16-in) bicycle wheel and then machining a
groove in the urethane that fit the cable radius (Figure 11).
Wi nch
The winch provided the mechanical recovery and hose/cable spooling
functions, and provided for continuous signal and power transfer through slip
rings. A rotary union permitted continuous water sample transfer when the
winch was in motion.
The original winch provided by InterOcean was powered by & variable
speed, direct current control motor through a planetary gear speed reducer
and a chain drive to the drum. The winch controllar converted alternating
current from the inverter to variable-direct current for speed control of the
drum. The hose/cable was secured to the drum with a Kellem grip, and the hose
member was attached to a rotary union in one end of the winch drum. A
20-conductor slip ring assembly extended outside the winch. The conductors
of the cable were hard-spliced to the slip rings inside the winch drum.
Changing slip rings in this configuration was a major task. In general, the
winch was too large to work around conveniently and was inadequately braced.
It also had no provision for manual operation in the event of an electrical
power failure, a serious safety consideration.
Between the first and second field years a new winch was designed, and
three copies were fabricated (Figure 12) that overcame the operational
shortcomings of the original winch. Two of the new winches had 100-meter
capacity drums and one had a 200-meter drum. The slip rings v/ere recessed
into the winch drum and a connector was added between the rings and cable;
this connector was accessed through a service port in the side of the drum.
This winch was powered by a control motor similar to the original winch, but a
worm gear speed-reducer was used. A manual crank (Figure 12) was included to
retrieve the cable in the event of a power failure and a manual brake was
added to the drum to prevent free-wheel ing in the event of a broken drive
chain. A slipclutch was incorporated in the drive assembly to prevent damage
to the air frame at the boom attachment points if the sensor/sampler were
21
-------
^
-: •* 1>6 -, ^.C.^'is-t:&a&&?t^^
-imtef, rr~l
Attachment Points
MljTo Helicopter Air Frame ft
V"*'' ^H^'''-'?'V?''^'''-V^^' ^iC^^-'Y^T^rnTT^^'^^^^^^^^^^'^*^^ ,8'>.'«»VJ>»«,'« iMjfcii"1**"*V —^ T^ __« _^
£ka^3^s£as^^a^£Sa^
Figure 10 Hinged
/ /
-------
38**- .I^Ag^¥iefe-^^V.&^«y5'5'J:-;
?£?&# afeA^
s SB:-;
I :/m S'
f
Figure 11 Lights^ S
/ ili
-------
Removeable Emergency Hand Crank
^»feij,^A.sas^i^
Figure 12 New 100-meter
-------
accidentally snagged on some submerged object. The chief advantages gained
from the redesign of the boom and winch assembly were in operational safety
and size reduction. The latter allowed relocation in the aircraft; the gain
in working space made it possible to replace the rear seat with an auxiliary
fuel tank mounted athwart ship. The tank doubled as a seat and extended the
operational radius of the helicopter.
ONBOARD ELECTRONICS
The onboard electronics included the console for the in-situ sensor
subsystem, the pump and winch controls, the graphical displays, the echo
sounder display and the power supplies (Figure 13).
Mounting Rack
An open-frame 19-inch relay rack was fabricated from aliniimim angle-
stock. The base was 20.5 inches (52 cm) square and was mounted on shock
mounts specially designed for helicopter use. The overall height of the rack
and mounts was 52 inches (132 cm). The back of the rack was shaped to fit the
contours of ttie helicopter cabin, and the rack was bolted to the deck of the
helicopter. The onboard electronics systems were installed with the static
inverter, the heaviest component (82 kg, 130 Ibs), at the bottom; then the
pump and winch controllers and the in-situ sensor console; and finally the
four graphical recorders at the top (Figure 13). The echo sounder display
was mounted on the side of the "relay rack."
Console
The console and digital-display module for the InterOcean model 514A was
provided in a box for portability. This was remounted in the relay rack to
consolidate the onboard electronic systems. The console operates on 12-vclts
d.c. power and provides the power to the analog circuitry for the sensors. It
also contains the analog to digital conversion circuitry, analog panel meters,
and a switchable digital display for greater resolution of the measurements.
This was the single most trouble-free component in the entire system. The
only change made in this component was a new front panel provided by
InterOcean to fit a 19-inch relay rack.
Graphical Displays
The original graphical recorder was a x,y,y', y" flatbed recorder with a
plotting area 27.9 cm (11 in) by 43 crn (17 in). Neither the electronics nor
the mechanical pencarriage assembly could withstand rotary-winged aircraft
vibration. Four separate strip chart recorders, Hewlett-Packard model 680A,
were substituted. An event marker feature was used to key regular depth
intervals to correlate the values noted to specific depths. Electric-writing
recorders were selected to avoid the inevitable untidiness of filling ink
reservoirs in an aircraft.
25
-------
m^Styff'fi'iv-^^'i-^i':.' '"..• •'"•
Strip Chart
Recorders(4)
^Mounting
Rack
eadout
Console
•^x.rt'Ksn
• — JL-i-r- •*- -'V
r-j-*. .-I*..!* •<;•*.•y.-.i.' ".
1. •••-/'.:.'•".-"^r»'uLifl5?-wA.*
^3S^^^ffifif »Wtf^^ S^*->--^Wffiifi3i^i^ 9^(D 9X
^easeajim^a^i^] %-^f ••'T^J^-*?~^=^^^
www«w-ii*iJSf*.we,!B :lr4itM*«^fi'^ssMt^jp^^mfr--<^^
•X\\ . .-i^X'T .Vf"1-"^.-'. V<^ '^ V-* ^, ."»"'v""| l rS
'-^v,;. Sa\k4fi5*'si ^ •- v".
'; ]
S>S^is-i*;w».-« w g'-^g^^hjfeBlii^gi^^g^^^^^^ataa^a
££^i ;j«>\«i.**^ ^^A^tN^-f^^'^^^^^iijfcv^'-1'^ h.-ffr' >'•'" ^rV^J^-^^-y^-^'-lAr^(^-t^w''''*e*fc-"*'::='Tr^•^*^j^.tA:Jgv«titJ^Maa-Haad«aH'iihtrfajM.Afcrfntii
V; Figure 13
^x.<- ^
-------
Winch Controller
The winch controller, Browning model M, was modified mechanically to
mount behind a relay-rack panel. The controller is powered by 117-volt
alternating current and provides a variable d.c. voltage to the winch notor
for speed and direction control.
Pump Controller
The pump controller includes the main power switch for the pump, the
"power factor" correction capacitor, a step-up transformer (117 V a.c. to 230
V a.c.) to pov/er the motor from a 117-V a.c. source, the starting-winding
capacitor, and starting-winding disconnect relay. The transformer and
correction capacitor are mounted on a panel on the rear side of the rack. The
remaining components comprise the standard "1/3 horse power" Berkley pump
controller and are mounted on the same panel as the winch controller.
Power Supplies
The power available on the aircraft is 23 V d.c. with maximum of 200
amperes available. Most of the sampling and measuring equipment requires 117
V a.c. power. During the first year of the NES, a 1,000-VA rotary inverter,
Leland Model MGE 37-400, was used to convert the 28 V d.c. to 117 V a.c. 60
Hertz. Rotary inverters function better than static inverters when used with
inductive loads, e.g., the pump motor; however, they are inefficient when
little current is being drawn from them. One thousand volt-amperes was just
sufficient to operate the winch and pump simultaneously. When the pump
clogged, the motor would draw more current than the inverter could deliver.
This repeated overloading of the inverter caused numerous inverter failures.
Larger capacity rotary inverters ware not available. After the first year of
NES, the rotary inverter was replaced with a 2,000-VA static inverter, Um'tron
model PS 67-324-1. The inductive load of the pump motor was corrected to a
power factor of about one by adding an appropriately sized capacitor in
parallel with the load. The winch motor required no correction since the
power supplied to the inductive load by the winch controller was d.c. The
static inverter is much more efficient than the rotary inverter, especially
during low load conditions. However, it weighs 54 kg (120 Ibs) more than the
rotary inverter.
The static inverter provided 117 V a.c. power to operate the pump, the
winch, and four strip chart recorders. The echo sounder and the console for
the in-situ sensors required 12-volt d.c. power. A Narco 28' to 12-V d.c.
power converter was used to power these components. The converter was
mounted on the back panel of the rack next to the pump step-up transformer.
Echo Sounder
A number of different pleasure-boat echo sounders of the revolving-
disk/flashing-light type have been installed on the helicopter since 1972.
Some of the manufacturers and models are as follows:
27
-------
Columbian Hydrosonics Model C8-324 Ray Jefferson Model 520
Heath Company Model MI-1031
However, none was found to be rugged enough to endure constant use and
helicopter vibration. The bearings supporting the revolving disk and motor
armature typically failed after a few months of field use and replacement
parts were not readily available.
The indicator unit was mounted on the front of the winch/pump controller
panel. The transducer is addressed in "Other Underwater Components."
SAMPLING MANIFOLD AMD SAMPLE STORAGE
Sample water delivered from the rotary-union on the winch was transferred
to a cylindrical clear-acrylic manifold for sample delivery. The manifold was
51 mm (2 in) diameter, 305 mm (12 in) long and was oriented horizontally.
Three sample ports on the bottom were fitted with plastic valves for sample
withdrawal. The supp-ly hose from the winch was introduced at one end and an
overflow hose was attached to the other end. A small stainless steel sink 102
mm (4 in) by 205 mm (12 in) was mounted below the manifold. The sink drain
and maniFold-overflow hoses drained outside of the helicopter. Both the
manifold and the sink were mounted on the side of a specially fabricated
aluminum sample-storage cabinet.
Samples were collected in 125-ml (4-oz) polyethylene bottles (flalge
£2003-0004) which were carried in plastic coated wire bottle racks, 48 bottles
per rack (Cole Farmer #6049-10). The sample-storage cabinet was constructed
such( that three racks of bottles fit on each of four shelves. A large space
at the bottom of the cabinet was reserved for tools, spare system parts etc.
OTHER UNDERWATER COMPONENTS
Solar Illuminance Meter and Secchi Disk
At the outset of NES, secchi disk measurements represented the sole basis
for estimating photic zone depth. Beginning in the second year, a solar
illuminance meter was used in addition to the secchi disk. A Montadero-
Whitney model LMT-8a was chosen. It functioned well during the remaining
years of NES and on other later applications. This instrument was deployed
independently of the sensor/pump system. A possible improvement would be to
integrate it with the sensor/pump assembly. This would have required that
only one system be put in the water and would have expedited the on-site
operation. Both the secchi disk and the solar illuminance meter readings,
when taken from the helicopter, were shallower than when taken from a boat at
the same site. This is, in part, attributed to reflections and refractions
at the v/ater surface from the rotor-wash turbulence. In the case of the
secchi disk, the demonstrated shading effect of the rotor blades also
contributes to the shallower readings. Generally, the 1-percent light level,
as measured from the helicopter float with the light meter, was about 2.2
times the secchi disk depth measured from th-a same position.
28
-------
Sediment Sampler
Bottom sediments were collected for nutrient and heavy-metals analyses.
The original sampler was a Phleger corer. The intent was to describe the
stratification of the nutrient accumulation in the sediments. The corer
resembles a bomb, with a core tube and cutter on its anterior. This assembly
is dropped into the water and rapidly decends to penetrate the bottom. When
it is extracted it recovers a 35-mm (1.375-in) diameter cylinder of sediment,
essentially undisturbed.
The corer was too heavy to recover by hand and interferred with the
primary mission of collecting water samples. Recovery after complete
penetration into the sediments sometimes resulted in failure to the boom
system. The concept of collecting intact cores was waived for expediency and
a small, light-weight clam-shell dredge, which could be deployed and recovered
by hand with a 3-mm (1/8-in) nylon line, was used for the remainder of the
study.
Echo Sounder Transducer
The echo-sounder transducer was mounted on a fixed bracket on the left
helicopter float. The transducer was positioned such that when the aircraft
was on the ground the transducer was about 20 cm (8 in) above the ground and
when the aircraft was on the water the transducer was submerged about 5 cm (2
in).
When different manufacturers' echo sounders were used, the transducer
was also changed to ensure compatibility.
29
-------
SAMPLING OPERATIONS
As a safety precaution, sampling should never be conducted with fewer
than three persons. Thejjilot, riding in the right-hand front seat, is
constrained to remain seated and to maintain control of the aircraft at all
times, in the air, on the water, and on land, as long as the main rotor is in
motion. If someone sampling has an accident or falls overboard, the pilot is
not at liberty to leave his place to lend assistance. The limnologist rides
in the left front seat and identifies the sampling sites. The sampling
technician rides in the rear of the aircraft. The limnoTogist records
observations concerning the site location, general water appearance, and any
other observations pertinent to the mission on a field data sheet. After
the sampling site is reached and the aircraft has landed on the water, the
technician debarks to the right float and deploys a station reference buoy for
the pilot. While on the water with the helicopter at flight idle for
sampling, the pilot maintains position on the site against wind and/or
currents.
_The limnologist, meanwhile, debarks to the left float and deploys the
in-situ sensor/sampler package. The limnologist then occupies the techni-
cian's seat at the system console to operate the winch and pump during the
cast" (Figure 3). The technician takes secchi disk observations and
bucket-dipped surface-water samples. The limnologist lowers the sensor/pump
package slowly through the water column until it contacts the bottom {or the
cable end is approached). During this process, the four analog strip-chart
recorders are activated to generate analog profiles of conductivity,
temperature, dissolved oxygen and turbidity. The sensor/pump is then raised
free of the bottom before the pump is activated, to avoid damage to the pump
from sediments entering the intake. Water samples are collected and preserved
as necessary by the technician. Samples are collected from a manifold inside
the aircraft. Sufficient time must elapse after starting the pump for the
hose/cable to be purged of water from the previous sampling (purge times were
routinely measured for each system). Sampling depths for the collection of
other water samples are chosen by the limnologist after inspection of the
analog profiles of the four parameters. The digital display on the console is
read and recorded on the field data sheets for all of the parameters at each
sampling depth. Upon completion of sampling near the bottom, the sensor is
raised to the next level, digital values are recorded, the hose is purged and
water samples are pumped. This process is repeated at each depth selected for
collection of water samples at a given site.
Integrated samples for some analyses are collected by continuing to pump
while raising or lowering the sensor package. Water collection is timed to
provide a uniform mixture of water over the integration depth range.
30
-------
Samples are collected in polyethylene bottles for all required types of
analyses, both chemical and biological, are preserved as necessary, and
stored in racks in the sample storage cabinet or in an ice-chest as sample
requirements dictate.
When sampling in rivers with noticeable current, as was sometimes the
case in both NES and the Atc'nafalaya Study, the sensor/pump package is not
lowered through the water column, but rather is used for surface samples and
measurements only.
At the conclusion of sampling, the limnologist retrieves the sensor/
sampler from the water, secures it in the aircraft and returns to the left
front seat before take-off. The technician recovers the buoy and returns to
his seat for take-off. Because of the high noise levels, headsets connected
to the helicopter's intercom system are used for communications while sampling
as well as during flight.
31
-------
CALIBRATION AND MAINTENANCE
The sensor analog electronics proved to be quite stable; however, the
wet/dry cycling of the sensors and the physical abuse from deployment/
recovery and sensor contamination necessitate daily calibration. During
routine operations the system is typically calibrated to standards and buffers
in the evening {Table 2) and checked against these standards the following
morning before flight. Differences between the system and standards for the
morning check are recorded and added/subtracted from the observations made
that day. The operational checks (Table 3) are performed both evening and
morning concurrent with calibration and calibration check.
Before beginning the day's operation the sampling technician reviews a
check-list for the presence of sampling supplies: bottles, reagents, ice and
other preservatives; removable sampling gear (sampling bucket, secchi disk
sediment sampler and tool kit); and spare parts (pump, pump rollers and pump
motor).
Maintenance that was accomplished outside of this daily schedule was in
response to catastrophic failures. In order to ensure reliable calibrations
and timely maintenance, it is necessary to keep a calibration/maintenance
person available to the system daily while the system is in use. This
calibration/maintenance person should be a qualified electronics technician
with an understanding of water chemistry and water quality sensors.
A portable artificial water column for calibration of the echo-sounder
was devised and is illustrated in Appendix C.
A buffer-board was designed to be added to the console to simplify
calibration. This modification was never fully tested or implemented; it is
illustrated in Appendix D.
32
-------
TABLE 2. DAILY CALIBRATION PROCEDURE FOR HELICOPTER BORNE WATER QUALITY
MONITORING PACKAGE
Parameter or Circuit
Procedure
Bipolar regulated power
supply
Ground current
regulator
Temperature Analog
zero
gain
Conductivity Analog
zero
gain
Depth Analog
zero
gain
Turbidity
Using a 4 1/2-digit digital volt meter, the positive
and negative outputs of regulator are calibrated to
8.000 V d.c. ± 0.001 V d.c.
Using a 4 1/2-digit digital volt meter, the voltage
between the two white test points is checked. The
voltage should be 0.000 V d.c. for proper function.
Using an ice water bath on the temperature sensor
with constant stirring and monitored by laboratory
thermometer, the zero pot is adjusted to yield
an electrical analog (0.100 V d.c. = 1.0°C) of the
thermometer indication.
An ambient temperature water bath monitored by a
laboratory thermometer is applied to the temperature
sensor. The gain pot is adjusted to yield an
electrical analog (0.100 V d.c. = 1.0°C) of the
thermometer indication.
With the conductivity sensor clean and dry, the zero
pot is adjusted to yield an electrical analog
of zero for conductivity 1.000 V d.c. = 1,000 ymhos.
With the buffer box provided by the manufacturer
set to 2,400 vimho position and the wire looped
through the sensor, the gain pot is adjusted to yield
2,400 ymho analog or 2.400 V d.c.
With the sensor connected, but not in water, the zero
pot is set to yield 0.000 V d.c. output.
With shunt points shorted on depth board the gain
pot is adjusted to yield the shunt value for that
serial number transducer from calibration sheet.
The windows are cleaned, the lamp current set to
60.00 and the full scale air calibration is set to
96.00%; then, with the optical system occljded with
a card, the zero is checked. , ... .*
(continued)
33
-------
TABLE 2. (Continued)
Parameter or Circuit
Procedure
Dissolved oxygen
zero
gain
The sensor is bathed in a stream of dry nitrogen or
dry helium and the zero pot is set to yield a zero
analog on the sensor console.
In ambient air, the gain pot is set to the 02
saturation value at ambient temperature on the
sensor console.
zero
gain
Echo sounder
With the sensor in a pH 7 buffer, the zero pot is
adjusted to yield 7.00.
Then with the sensor pH in a 10 buffer, the gain pot
is adjusted to yield a pH 10.00 on the console
display. With the sensor in pH 4 buffer, the gain
adjustment is checked for pH 4 on the display.
With the sensor in a 10'x4" water column (Appendix C)
the speed of the motor is adjusted to yield blips at
exactly 10 foot intervals after the motor is
lubricated. The oscillator and electronics alignment
are checked by seeing at least 10 blips at 10-foot
intervals or to 100 ft in a 10 ft pipe.
34
-------
TABLE 3. DAILY OPERATIONAL CHECKS OF THE IN-SITU SENSOR/SAMPLING PACKAGE
Equipment
Procedure
Strip chart recorders (4 ea)
Static Inverter (output)
Winch
Pump
With the input shorted, the zero control is
set to position the pens to zero on
the paper, for each recorder.
The inverter is turned-on, the winch brake is
set, and the winch started slowly while the
output voltage, frequency and current are
monitored via the meters on the inverter.
The inverter should deliver 10 amps to the
winch without a voltage or frequency change.
After the inverter check, the winch drive is
then checked in both the deploy and recovery
directions. The winch drum slip clutch
tension is checked.
The sampling pump is checked for worn rollers
and impeller, worn motor and pump bearings
and hard starting of the punp motor (This is
only accomplished at the evening check,
except for the ease of starting).
35
-------
SAMPLING CAPABILITIES
The airborne sampling system is totally integrated-, the equipment has
been custom-tailored to fit the interior contours of a UH-1H model "Huey"
helicopter; thus the helicopter platform has become part of the system.
The helicopter has an operational period of about 2 hours at 100 knots
(airspeed limited by floats), without auxiliary tanks. An accessory auxiliary
tank can extend that period to about 3 hours.
It should be noted that 3 hours times 100 knots air speed does not
necessarily equal 300 nautical miles of ground track, because of head winds
or tail winds. Fuel consumption is cut in half when the aircraft is at flight
idle during sampling. Travel time from the operations base to the first
sampling site and the distance (or time) between sampling sites strongly
influence the number of sites that can be covered between re fuel ings. Factors
of lesser influence include the type of sampling, the number of samples par-
site and the depth of water at each site. Thus it is difficult to make any
meaningful statement about the number of sites that can be sampled in any
given time period. However, during the NES, the sampling extremes ranged from
one site per refueling (far away from the base) to nine sites per refueling
when the sites were close to one another, and to the operations base and the
sites were shallow (less than 10 meters).
The crew required for normal sampling operations consists of a pilot and
two sampling personnel. The aircraft can accommodate two additional sampling
personnel or observers. The three sampling systems are identical except for
the sizes of the winches and the maximum sampling-depth capabilities. Two
systems can sample to a maximum depth of 80 meters (263 ft) and the third
system can sample to a maximum depth of 172 meters (555 ft).
The system is capable of in-situ measurements of: temperature, conduc-
tivity, dissolved oxygen, pH and turbidity (Table 4).
The water-sample cabinet can hold 576 125-ml (4-oz) samples in storage
racks. Additional samples may be stored in other locations within the air
craft, as necessary. Water samples of any volume can be collected at a rate
of 7 1pm from any depth up to the maximum. The sampling pump can be operated
while the winch is retrieving the sampling assembly, in order to collect a
depth-integrated sample of the water column. The total number of samples
that can be collected is limited by the storage space in the aircraft. The
water samples collected may be used for a variety of water quality analyses
from water chemistry to phytoplankton analyses. If the pump and hose/cable
36
-------
TABLE 4. SPECIFICATIONS OF THE IN-SITU WATER QUALITY MONITORING PACKAGE
Parameter
Range
Precision*
Conductivity
Temperature
Depth
DO
pH
Turbidity
0 to 65,000 ymhos
-5 to 45°C
0 to 100 m
0 to 20 mg/1
2 to 12 pH
0 to 100% transmission
±10 ymhos
± 0.02°C
± 0.3 m
±0.2 mg/1
± 0.05 pH
± 2% transmission
* From the manufacturer's specifications.
were replaced with a Teflon pump housing and Teflon hose in the hose/cable,
the system would be satisfactory to sample water for pesticides.
The system has most comnonly been used to sample in lakes and other non-
flowing waters. However, it has also been used to collect surface data and
water samples in rapidly flowing streams, e.g., some Atchafalaya study sites.
In this case, the in-situ package was not lowered through the water column,
because difficulties in maintaining orientation of the helicopter in flowing
waters enhance the hazard of snagging the sensor package or fouling the pump
system.
Bottom sediment samples can be collected with a small, hand-operated,
clamshell-type dredge. More sophisticated sediment samplers have been used
from this sampling platform, e.g., cores using a Pflegar Corer, but there is
not sufficient room in the cabin to do this type of sampling without removal
of the water quality sampling hardware. As we used it, the Pflegar Corer
also required a cut-out in the floor of the craft and removal of the athwart-
ship auxiliary fuel tank.
37
-------
DISCUSSION
Originally, three identical, off-the-shelf, in-situ water-quality
monitoring/sampling systems v/ere procured. Throughout the 4 field years of
NES, for an additional 2 field years on Atchafalaya Basin Water Management
Study, and during other subsequent studies, the systems v/ere constantly being
modified to better fulfill our requirements.
Helicopters are unique and versatile sampling platforms, however, they
do have certain limitations. Space and weight capacity for samples and
equipment, and electrical power for sampling hardware are usually limited
aboard a helicopter. Mechanical vibration is usually quite severe. Light-
weight, compact sampling and measuring hardware helps to reduce the hardware
load and maximize available working space for sampling personnel. Sample-
storage containers that are designed to optimize otherwise unusable space are
a considerable advantage. Safety considerations of personnel, sampling
hardware and the craft itself are best fulfilled by securing all equipment and
sample-storage containers during flight. Small, light-weight, overboard
sensors and samplers increase the speed of sampling operations. During the
first year the system was in use, when frequent clogging of the original pump
impaired our sampling efforts, v/e considered changing to a "bottle type" water
sampler. At that tine we concluded that there was neither the space to carry
five or six bottla samplers in the aircraft nor tine to make the nultiple
casts at a single station required to sample from several different depths
with a single bottle sampler. When sampling from a helicopte^ to maximum
depths of about 170 meters, the pump-sampling system is the desirable
technique. It is believed that the maximum depth that would be practical to
sample with a pump-sampling system in a Huey is probably about 350 meters.
Sampling from greater depths with a pumping system would require a larger
winch than would be practical in this aircraft.
Electrical power is limited on helicopters and is usually 12 or 24 volts
d.c. It is more efficient to operate the sampling hardware directly from the
aircraft power, whenever possible, to minimize the amount of power transformed
to other voltages. The exception to this is when a large amount of pcwer is
required remote from the craft, e.g., a submerged electric pump. The power
dissipated in the wires becomes significant when the wires are long and small,
or when the currents are high (greater than 1A). In this case the power
should be converted to a.c. and transformed to some higher voltage. The
remote equipment should be selected for high voltage, low current require-
ments. Certain types of equipment require 117 V a.c. power (e.g., drive
motors in some strip chart recorders). With such equipment, an inverter must
be used to convert aircraft power to 117 V a.c. 60 hertz. Some inverters have
poor or non-existent filtering in the input circuit which results in the
38
-------
reflection of electrical noise back into the aircraft power bus. This
power!ine noise and other electrical noises indigenous to aircraft can mask
the data signals in the measuring equipment and possibly lead to generation of
erroneous data. Dissolved oxygen and pH sensors and circuitry are particu-
larly susceptible to electrical noise interferences and ground loops. If
multiple pieces of equipment are to be used in the water simultaneously, they
may also exhibit some interaction with one another.
Vibration levels in aircraft, particularly rotor-winged craft, are high
and potentially damaging to electronic hardware and water quality sensors,
especially the very fragile pH electrodes. Shock-mounting of all stationary
components and cushioned restraining cradles for on-board storage of all
deployable systems will reduce vibration damage.
The sensor/sampling systems on the whole functioned quite well; however,
they do require constant attention in terms of calibration and maintenance.
The dissolved oxygen and pH sensors are those rcost vulnerable to damage and
drift from abuse.
One problem, quite unique to helicopter water-sampling platforms was
discovered. The downwash from the rotor blades of the helicopter physically
disturbs the water's surface layers. A brief dye experiment conducted at
Lake Mead (Figure 14) indicated the disturbance extended to at least 1 meter
depth directly below the craft. The upper 20 to 50 centimeters move outward
radially from the landing site and that water is replaced by upwelling water
from below. While the full extent of this effect is not known, it is apparent
that all structure in the upper meter of the water column is destroyed.
The helicopter has been demonstrated as a cost-effective sampling plat-
form when many sites must be sampled and the travel time between sites is
excessive in relation to the time required for actual sampling.
39
-------
•, . i •..••ivii...'ii.'.-,-" •:.••;••• ';. .i- , •
::::
Figure-14 Dye^ludy of jleVlcopter^otorwash-^urbulence /ffects on Lake Mead.
-------
APPENDIX A
THE DISSOLVED OXYGEN CIRCUIT
The dissolved oxygen sensor produced by Yellow Springs Instrument Company
functions as a variable-current source when polarized with a 0.7 V d.c.
potential; the current is directly proportional to both the partial pressure
of the oxygen and the temperature of the membrane and the sensor assembly.
The sensor current is converted to an analog voltage by the operational
amplifier (op-amp) (AD 523 L) and the feed-back resistor network (560 K and
250 K trimpot). The sensor current and the output of the "current to voltage"
transducer are zero in a zero mg/1 oxygen environment. The op-amp output
voltage, Vo, is calculated by Vo = -I x Rf.
The analog multiplier (AD 532 S) has differential inputs for both "X"
and "Y" and its output voltage is Vo = (X1-X2)(Y1-Y2)/10. The output of the
op-amp, a positive voltage, is connected to the negative "X" input of the
multiplier. The positive "X" input is grounded. The temperature analog
output (T), from the temperature board is connected to the positive "Y" input.
The negative "Y" input is connected to a constant +3.25 v d.c. This voltage
is developed by the adjustable voltage divider connected to pin 10 (-"Y"
input). The voUage-te-Tiperature relationship is 0.1 V d.c. = 1.0°C. The
algebraic representation of the complete circuit is DO (0.1 V = 1.0 mg/1) = -I
x 0.67 (T - 3.25J/10. This is a linear approximation of the second degree
polynomial DO = 0.7945 I/(0.0104 T2 + 0.2045 T + 3.7879) that approximates the
sensor performance between 0° and 25°C. The linear approximation was adjusted
to minimize differences (±2%) from the polynomial in the temperature range of
6°C to 19°C.
The MC 1568 C is a voltage regulator circuit providing bipolar (+ & -)
15 V d.c. required for the multiplier. Other modules in the sensor package
operate on the original bipolar (+ & -) regulated 8 V d.c. The voltage
divider from the negative supply, the 33/K and 1.6/K resistors to ground,
provides the polarizing voltage for the sensor.
Before calibrating the oxygen circuit, the temperature calibration should
be assured. To calibrate the circuit and sensor cc.Tibination, the following
items capabilities are required:
A 3^-digit digital voltmeter, a source of dry nitrogen or helium
gas, water in the 10°C to 15°C range that is nearly saturated with
oxygen, and the capacity for a Winkler titration for oxygen
determination. The sensor is first bathed in a stream of the gas
to exclude oxygen from the membrane. While this zero oxygen
environment is maintained at the membrane, the 10 K trimpot
41
-------
connected to pins 1, 4 and 5 of the op-amp (AD 523 L) is adjusted
to yield 0.000 V d.c. at tp-1. Next the 20 K trvnpot connected
to pins 2, 5 and 9 of the multiplier (AD 523 S) is adjusted to
yield 0.000 V d.c. at tp-2. At this time the sensor is removed
from the zero oxygen environment. The 10 K trimpot connected to
pin 10 of the multiplier is adjusted to yield +3.250 V d.c. at
tp-3. A sample of the water to be used for the instrument calibra-
tion is then evaluated for oxygen concentration using a Winkler
titration; the temperature and oxygen sensors are placed in the
remaining water. The water is gently stirred while the temperature
and oxygen sensors are allowed 2 minutes to reach equilibrium with
the water. Once equilibrium is achieved, the 250-K trimpot in the
"feedback" circuit of the op-amp (pins 2 and 6 of AD 523 L) is
adjusted to yield an analog of the dissolved oxygen concentration,
found from the Uinkler titration, at tp-2 (+0.1 V d.c. per 1 mg/1 of
oxygen). This process may be repeated to verify stability.
42
-------
15Vil c
TP
1
From Output of Temp Analog
5 G
33
AD532 S
1 Y1 •
10 YZ -
—oTP
100 K
•IGViJ c
Power Input
1fU/,l , „
Giul — » .
I
1
MC
15G8
C
6
t/vwv
& G
i
U
f
£
T°
-i-E
' o
o -1
0 ~
33 K 5
1 G K<
f
r:^. >.n>
~L
_2 y,,,[
TP
IbVd c
*- OUTPUT
•15Vclc -ISVdc
15Vt| c
0 7Vti c 10 Oxygen Sensor Shield
Figure A-l. The dissolved oxygen circuit.
-------
APPENDIX B
THE DIFFERENTIAL-INPUT pH-MEASURING CIRCUIT
The glass pH-measuring electrode and the Lazaran reference electrode
function as a variable voltage source with an extremely high internal
impedance (10+6 to 10+s ohms). The output potential of the pair is zero at pH
7.68; the voltage analog from the electrode pair is 59.2 millivolts per pH
unit. The temperature factor is 0.77 millivolts per degree to be sub-
tracted above 25°C and added below 25°C.
The high impedance of the sensor, pair requires the use of an amplifier
with a high-input-impedance. To minimize system susceptibility to electrical
noise and "ground loops," a high-input-impedance differential-buffer-amplifier
(HA2005) was placed in the sensor mounting block. The input leads from the
electrodes to the amplifier were about 1 cm long. The buffer amplifier has
an input impedance of IQiz ohms and unity voltage gain. The 100 K trimpot
connected to terminals 1 and 5 is the balance control and is adjusted for zero
output in a pH 7.00 buffer. The differential output of the buffer drives the
differential inpuc of the instrumentation amplifier (AD 520 K). The gain of
the device is controlled by the ratio of the 100 K resistor (pins 9 and 11)
and the 47 K resistor and 10 K trimpot (pins 5 and 7). This gain function
raises the analog from 59.2 millivolts per pH unit to 100 millivolts per pH
unit. The offset terminal (pin 12) is ulilized to shift the sensor zero
output voltage corresponding to a pH 7, to -0.7 volts or the 100 millivolt per
pH unit analog. The negative output is inverted by the following stage.
The output inverter stage (AD 741 K op-amp) accomplishes the temperature
correction of the sensor. The temperature correction is a linear additive
correction. The 0.77 mv/l°C at the sensor is the equivilant of 0.013 pH units
per degree. The zero point on the temperature correction curve is 25°':, and
0.013 pH units is added per degree Celsius below 25°C. The 1 K and two 76.8 K
resistors at the inverting input of the op-amp comprise a three-input adder.
The 1,000-ohm "feedback" resistor from the output to the inverting input
provides unity gain when the temperature is 25°C. The 100 K trimpot is set to
-2.5 V d.c., corresponding to -25°C, which is added through 76.8 K resistors
to the positive temperature analog. As the temperature analog falls below
25°C (+2.5 V d.c.), the analog temperature difference from 25°C is added to
the pH analog at the rate of 1,000/76.8 K per degree Celsius (0.013 pH/l°C).
This circuit is powered by the original regulated 8-volt supply.
44
-------
Before calibrating pH, the temperature calibration should be assured.
To calibrate the circuit, the following items and capabilities are required:
two Celsius thermometers with 0.2 degree or finer graduations which include
25°C in their range, a 3 1/2-digit digital voltmeter, pH 7 and 10 buffer
solutions, a 100-ml beaker, a 250-ml beaker, a 500-ml beaker, and a source of
25°C water for temperature control.
The pH 7 buffer is put in the 250-ml beaker which, in turn is put in a
water bath in the 500-ml beaker at 25°C. When the buffer is 25°C, the pH
electrode assembly is immersed in the buffer and the 100 K balance trimpot is
adjusted to yield 0.000 V d.c. between TP-1 and TP-2. All following voltage
measurements are made referenced to TP-7 (ground). The 10 K offset trimpot is
adjusted to yield +0.700 V d.c. at TP-3 and the 500-ohm zero trimpot is
adjusted to yield 0.700 V d.c. at TP-4. The 100 K temp-cal-comp trimpot is
adjusted to yield -2.5 V d.c. at TP-5. With 25°C water on the temperature
sensor, the 10 K zero trimpot is adjusted to yield +0.700 V d.c. at TP-6. The
pH electrodes are thoroughly rinsed and immersed in pH 10.18 buffer at 25°C,
and the 10 K gain trimpot is adjusted to yield +1.018 V d.c. at TP-6.
The process should be repeated to verify linearity. If the expected pH
to be measured is below pH 7 the pH 10 buffer should be replaced with pH 4
buffer and the gain trimpot adjusted to yield +0.400 V d.c. at TP-6.
45
-------
-is.
en
Input
From
Sensor
Prenmp
100
•BVtl c
8V.I c
SENSOR PREAMP
HA-2005
Figure B-l. The differential-input and pH-measuring circuit.
-------
APPENDIX C
WATER COLUMN FOR CALIBRATION OF ECHO SOUNDER
The calibration of the echo sounder posed a special problem. A water
column was necessary for depth calibration and for electronic alignment of
the driver and receiver circuitry. An artificial water column was fabricated
from a length of 4" A3S sewer pipe, a pipe cap and a plastic trash can. A
piece of 1/2-inch thick plastic was mounted at a 45° angle in the trash can
as a reflector. The 4-inch ABS pipe was cut to a length such that the path
length from the transducer to the 45° reflector and down the pipe to the pipe
cap was exactly 10 feet. The trash can/pipe assembly is placed under the
transducer with the pipe lying on the ground. The pipe and trash can are
filled with water to cover the transducer. Electronic alignment of the
circuitry is accomplished by adjusting for the greatest number of reflections,
or with an oscilloscope. Calibration is accomplished by adjusting the speed
of the revolving disk.
47
-------
10 Ft.
00
Trash Can
Echo Sounder Transducer
4" ABS Sewer Pipe
7
Pipe Cap
45° Reflector
Figure C-l. Ten-foot water column for calibration of the echo sounder on the helicopter
but out of the water.
-------
APPENDIX D
CALIBRATION BUFFER BOARD
Calibration of the original sensor package required disassembly of the
pressure case to access the calibration controls. A five-channel analog
buffer was designed and installed on one system to evaluate the performance of
the circuit. Evaluation was never completed. The circuit of a single buffer
amplifier is shown. The concept is to be able to offset the zero point ±0.5
volts and to vary the gain from 0.6 to 1.5. The five-channel buffer board
was mounted on the read-out console with the screwdriver-settable zero and
gain controls accessible from the front of the console. Power for the buffers
was taken from the read-out console. This circuit should reduce the need to
open the pressure case for calibration to a rate of only one calibration in
four.
49
-------
un
O
input from insitu package
10k
100k
•WVA/V -WWWWAA
•^ 10k
15Vd.c.
Output to
Console Input
-15V d.c.
Figure D-l. Schematic of one channel of calibration buffer amplifier.
-------
LITERATURE CITED
U.S. Environnental Protection Agency. 1974. national Eutrop'nication Survey
Methods for Lakes Sampled in 1972. National Eutrophication Survey
Working Paper No. 1. Environmental Monitoring and Support Laboratory,
Las Vegas, Nevada, and Environmental Research Laboratory, Corvallis,
Oregon. 40 pp.
U.S. Environmental Protection Agency. 1975. National Eutrophication Survey
Methods 1973-1976. National Eutrophication Survey Hor'r-.ing Paper No.
1975. Environmental Monitoring and Support Laboratory, Las Vegas,
Nevada, and Environmental Research Laboratory, Corvallis, Oregon. 91 pp.
51
-------
TECHNICAL REPORT DATA
(P'easereadI>'.s:r.'.c::o".soi: •! s reverse oetcrs -omp!e.'"-gi
= ORT"O
EPA-600/4-31-005
13 RECIPIENT'S ACCESSION 'J
ITLE -WO S'_'3T:TL =
DEVELOPMENT OF A HELICOPTER V.'ATEP QUALITY
.'IONITORiNG/SAMPLIMS SYSTEM
5 REPORT DATE
February 1931
ORiiANIZATiG.V CO^r
8 PEHrORMI.\S ORGAi\!ZATIOiii R=?C
H. Michael Lov/ry
9 PERFORMING GRGANiZATION NAME AND ADOR2SS
Environmental Monitoring Systems Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Las Vegas, NV 89114
10 PROGRAM ELEMENT NO
11 CONTRACT, OR-,NT .\O
s
12 SPONSORING AGENCY NAME A.\O ADOnESS
U.S. Environmental Protection Agency--Las Vegas, NV
Office of Research and Development
Environmental Monitoring Systems Laboratory
Las Vegas, NV 89114
I 12 TV?E CF
T AND PERIOD COVEflcO '•
14 S
15 SUP°LE.V1ENTAHY \'OTHS
AGfc'ICv COOE
EPA/ 600/0 7
16 ABSTRACT
This report describes the helicopter water quality sampling system developed
for use by the Nacional Eotrophication Survey and subsequently used in support'
of other water quality projects. It also describes the salient oroblens
associated vrlththe use of a helicopter as a sampling plat^crm and t^e iiudifications '
medejn the system to fulfill cur needs. This system is useful to other groups '
involved in sampling large numbers of water quality sites distributed over large
or inaccessible areas.
17 KEY V/ORD3 AND DOCUMENT ANALYSIS
a DESCRIPTORS
Ai rcraf t
Helicopter
Water pollution
3 6lSTPI3UTiCiV S rATEMENT
RELEASE TO PUBLIC
b IDENIIrlEHS/OPE.Vi SNCED TE.R.V.S
National Eutrophi cation
Survey
Airborne platforms
19 SECURITY Cu/SS , H.is Rfac.-:)
UNCLASSIFIED
20 SECUS!TV CLASS 'T- 'j pceel
UNCLASSIFIED "
o COSATI I icId'C'CJ^
48C
68D
j
j
?.' NO Or PAGES
^ 61
EPA Form 222C-! (Re- :-77!
ED! TIOM i s O33O1.ET e
------- |