Datalogging water lines


Thread Starter

Pierre Desrochers

Hi all -

As usual I have the highest confidence on this lists members to suggest some ways to accomplish this new task. Thanks all in advance.

We have some retractable Thermal Massflows installed on various points of a grid of water conduits. Those conduits have I.D. of 3 to 108 inches and insertion fitting are 1inch Fem. NPT.

The flowmeters are giving 4/20mA outputs. We need :

1. Log average flow for periods of...Ex: 10s every 15 min. during a period of 100 days...

2. Retrieve those Data and visualise them on an Excel spreadsheet

3. Draw graphics and print them...

All this is not much of a challenge but :

A . Instruments are located in manholes on busy streets
B. All instruments must be battery operated.
C. All electronics must be -20C to + 40C
D. We want to retreive the data with a laptop when we pass over the manhole with a service vehicle ... (could stop for a few minutes...)
E. The battery should be an ordinairy lead/acid type (like golf carts, etc)

Your help will be greatly apreciated.

Pierre Desrochers
System Integrator
Integral Instrumentation Inc
[email protected]

Four methods are available for filtering of spreadsheet data.

1. Moving Average
Calculates according to the formula:

y(i)=1/(2k+1) S x(j)

where the sum is from j=i-k to i+k
x(i): input value at time i
y(i): calculated output value at time i
k: one-side lag; the total lag is 2k+1

2. Gaussian Filter
Calculates according to the formula:

y(i)=S a(j) y(i+j)

where the sum is from j=i-k to i+k
X(i): input value at time i
Y(i): calculated output value at time i
a(j)= 1/[root(2p) s] exp[-j^2/(2g^2)]
.s : the standard deviation or width parameter, defined by
f: cutoff frequency

3. Exponential Filter
Calculates according to the formula:

y(i)=y(i-1)+(1-f) (x(i-1)-y(t-1))

X(i): input value at time i

Y(i): calculated output value at time i
f: damping factor

4. Point filtreing: For details, if interested
contact me [email protected]
(It takes some time to type)

The above is only part of your answer:
Battery operated and reading from distance,
I will skeep that one. I'm sure you can help
Two more points to be discussed:
1. The exact point of insertion is important.
I will supply the formula if required.
2. Excel will be adequate up to the maximum size
of the spread sheet. Then you can manipulate
the columns and finaly obtain the graph or
bar chart (up to you).
In the event the spread sheet size is adequate
but you don't like the Excel visual,
then you might use Origin 6.0.
This math/graphic package is exceptional in
performance and quality/versatility of
graphics. It accepts data from Excel spread
sheet + so many more, as well as andvanced
math package. Origin was designed having
Excel in mind.

Now: if the size of the spread sheet Excel is
insufficient, then you coud use Mathcad 8
Professional. In that one the maximum size
of array me be managed so that its limit is
the limit of your computer.

You may also check with Natiopnal Instruments.

Note: Origin 6.0 is produced by Microcal.
You can contact Bob Sullivan
<[email protected]>
Refer my name is perfectly OK.

This reply from: [email protected]

Anthony Kerstens

Just a thought, but if it's down a manhole, would
it also need to be explosion proof just in case of
underground gas line or sewer vapors inadvertently
leak in?

Anthony Kerstens P.Eng.

Pierre Desrochers

At this time no mention of anything like this was made by our client but interestingly your suggestion is very wyse... I will bring it up at our next meeting. TY


Robert Lockert


We faced a similar challenge about 6 years ago. The purpose was to monitor peak flow rates. We chose to log the total flow for every minute, every day. We used Control Microsystems RTU's at each location.

Generally we found the use of pulse meter heads to be more effective than integrating a 4-20ma. signal. Most 4-20ma. transmitters don't work below
10-11 volts so you really need a 24 volt supply (ie 2 deep cycle batteries). Battery maintenance became the greatest problem (and expense) so eventually utility power was brought in where practical. Road side solar panels were used where utility power source was too distant.

In your case, with a 10 sec. sample every 15 min. you should power the loop from the RTU and run the current through a relay that you can turn on and off before and after your sampling period. Also try to use RTU and modem sleep modes if possible to conserve battery power.

We wrote a custom app. that polled all locations through a 900mhz modem (radio) set up as a master. Each datalogger included a slave modem that uploaded requested data when the master was nearby and able to communicate.

Real time clocks will drift. If one location needs to be compared or added to another you must have a mechanism to periodically resync the clocks. (Ideally at download time)

I'd suggest storing the data in some format other than Excel, then import the file to Excel for your specific needs. File size and speed are the issues.

Our experience was that once the customer had a little data, he wanted more and more frequently. That need soon justified the installation of dial up phone lines to permit overnight retrieval from each location. Furthermore any location could be called on demand to view current flows, pressures etc.

Costs are about the same if you consider that a good spread spectrum modem will likely cost more than a 300/1200 bps modem and the line installation combined. The data collection costs of putting a man in a truck more than once a month are about the same as the monthly line costs.

A very viable option today is cell phone. Check signal strength at each location (undergrond) and try to call out with a laptop before committing
to this though. I'm not too sure about their acceptability at -20C either. These low cost personal communicators are interesting too but again -20C?.

The only significant problem we've had over the years is ground water penetration at one particular location. An enclosure with suitable ingress protection (IP xxxx - don't recall the numbers now) for all equipment including flow transmitter and batteries might be needed.

Feel free to contact me directly if you have any specific questions.


Bob Lockert
Blocke Communications
Calgary, Alberta
[email protected]
(403) 286 6711