Professional Documents
Culture Documents
Issue 2.1
January 2000
Socrates Operating Manual
This operating manual documents the features and use of the Socrates analyser
together with the associated PC software Version V1.6
Primayer Limited reserves the right to amend any specification without notice. Primayer will not be
responsible for any consequential damages arising from such changes, or errors contained in any
publication.
Windows 95, 98 and NT are registered trademarks of the Microsoft Corporation. All other trademarks are
acknowledged.
Table of Contents
1.0 Introduction…......................................................................................................................... 2
3.0 Installation............................................................................................................................ 13
3.1 Connections.................................................................................................................. 13
3.2 Sealing.......................................................................................................................... 13
3.3 Use................................................................................................................................ 14
3.4 Battery replacement...................................................................................................... 15
1 Introduction
Socrates ™ is a new class of instrument for analysing night-flows and leakage in water distribution
systems. Socrates ™ uses novel, patented, autocorrelation techniques, plus artefact filtering, to carry out
high accuracy surveys of district metering areas (DMA’s) and waste areas.
• Operational troubleshooting on District Metering Area’s. With its diagnostic precision Socrates
is able to probe deep into night-flows to reveal the lowest baseline flows, pointing engineers towards
real areas of high leakage. This is very useful on DMA’s with trade users who use water
intermittently.
• Company specific measurement of legitimate night use on small District Metering Area’s.
Socrates shows baseline leakage with user events clearly superimposed. From this information
legitimate night use is calculated.
• More realistic estimation of leakage. Socrates analyses thousands of data values per night and
uses these to form a statistical model. Baseline leakage is deduced from this information.
Nightlines are the closest we get to measuring leakage but these always include significant amounts of
legitimate customer night use. Separating leakage from use is not always easy. Conventionally,
averaged nightline data is collected and from this a fixed estimate of night use is subtracted. However,
night use varies! Fixed values can never reflect the changing conditions that actually prevail on site.
Socrates solves the problem by extracting statistical pattern information from the nightline to dynamically
adjust ‘fixed’ night use estimates to suit changing site conditions.
Socrates collects thousands of data values per night. After removing measurement artefact, the data is
autocorrelated to find its micro-period. This is the shortest period over which values must be averaged to
remove extraneous hydraulic effects like oscillations and surges. This reduces the raw data to, typically,
500 micro-period values and these expose the crucial details of the night-line.
Minimum nightly micro-values vary about mean levels and a statistical estimate of the actual leakage
may be derived from the distribution over a series of nights
Socrates is suitable for use on District Metering Area’s (DMA’s) or Consumption Monitor Areas which
are supplied through a single meter; it should be fully isolated from neighbouring networks and there
should be no maintenance intervention during the analysis period.
.
2 Theory of Operation
Like all measurement systems, PIT suffers its own share of inherent errors, which we call measurement
artefact. Additionally, there are greater pitfalls arising from certain extraneous hydraulic phenomena.
Unless these are detected, simple PIT methods cannot be used with confidence.
This inherent variation, which is independent of flow, is known as “measurement artefact”. It occurs for
all types of meter, to a greater or lesser extent, and has to be accommodated when making
measurements at high resolution. A positive or negative correction must be applied to each
measurement to compensate for the variation.
Socrates calculates a set of coefficients representing how much on average each segment (i.e. the
difference in time between two successive pulses) is longer or shorter than the average for the entire
group, and adjusts the measured water volume accordingly. Typical correction factors are shown below,
and illustrate the variation which occurs in a typical pulse head.
The graph opposite shows such an effect; the normal Hannah More 2 Oct
68
Flow Time
flow pattern is distorted by a transient. It has been PU100A only HO212ft
Head (Metres)
found that the most reliable method of detecting flow 64
Flow (Litres/sec)
10
Socrates monitors the pressure record with a moving
fixed time window. A pressure transient is identified 5
if the difference between any two samples in the
window exceeds some threshold. The window length 0
Figure 4 Multiple
Clevedon flowDrv
Yeolands transients without associated pressure transients
24/6 YJ2412ft 31
(arrow indicates
PU100A + PULSEpoint
DISC of Socrates measured minimum).
Head (Metres)
30
29
2
28
Flow (Litres/sec)
1.5 27
1.0
0.5
47 48 49 50 51 52 53 54 55 56 57
Minutes from 0230
PRV systems can produce notable cyclic oscillations. The graph shows an oscillation with a period of 10
seconds. Cyclic activity is not confined to PRV-controlled networks; it can occur on open loop systems
because of frequency response effects. It is important to eliminate such effects, as the minimum values
obtained during such cycles will give erroneous data. What is required is a method of obtaining the
lowest possible mean flow measured over the shortest practicable time because, as was explained,
longer averaging times cause unrealistic errors.
25
Flow (Litres/sec)
Average Flow = 5 L/sec approx
Flow cycle repeats every 10 seconds
4
65 66
Minutes from 0230
Socrates uses autocorrelation to calculate optimum averaging times to reduce the effect of short-term
oscillations in night-flows. Autocorrelation is a powerful technique which reveals hidden cycles buried in
noisy data. (A related technique, cross-correlation, is used in leak noise correlators.) Socrates
autocorrelates to detect periodic hydraulic oscillations whilst rejecting other fluctuations caused by
users.
The graphs below show examples of accurate night-flow flow produced by Socrates.
The first graph (Figure 6) is from a PRV- The third example (Figure 8) is data from a
controlled site. The Socrates-derived minimum mixed urban/industrial site. Conventional
flow is 3.3 l/sec, whilst conventional measurements show a steady rise in baseline
measurement techniques produce a value of 4.0 from day 15 to day 21, followed by a dramatic
l/sec, some 25% higher. rise on day 22. Socrates-derived data, however,
shows that the true baseline activity has
remained much steadier; indeed, it has fallen on
Causeway 1 Oct to 15 Oct, 2543 props +0 Trade some days when conventional measurements
5 have shown a rise. The difference is due to
Litres/sec
Minimum 15 min average
industrial use of water. Comparing readings on
day 31 shows that the conventional figure is 5.3
4 l/sec whist Socrates has produced a value of
2.9 l/sec, thus the conventional figure is 87%
too high.
Socrates minimum
3
Pill Sept 6 to Oct 7, 2000 props + Trade
20
Litres/sec
Minimum 15 min average
2
1 6 11 16
15
Day
some 30%. 0
1 6 11 16 20 25 30 35
Day
Rocks Oct 2 to Nov 1, 881 props +0? Trade
5 Socrates minimum
0 5 10 15 20 25 30 35
Day
For the calculation of leakage and legitimate night use it is recommended that approximately 30 nights
data are analysed. The consumption monitor area should be supplied through a single meter, it should
be fully isolated from neighbouring networks and there should be no maintenance intervention during the
analysis period. For the purposes of calculating consumption, the monitor area should have a
reasonably stable rate of leakage and the general pattern of consumption should be fairly stable. For
use of Socrates the DMA should be supplied through a single meter, it should be fully isolated from
neighbouring networks and there should be no maintenance intervention during the analysis period.
The components of core flow comprise: leakage flow, continuous ‘use’, and plumbing losses.
Leakage flow is relatively stable over the analysis period. Continuous use, however, varies randomly – an
example being flow from a tap which has not been turned off properly; one night it may be left dripping
rapidly, the next night it is fully off. Plumbing losses (due to cistern overflows, faulty tap washers etc.)
also happen randomly. It is these effects that cause the nightly variation in core flow. Figure 10
illustrates these components.
To summarise - core flow levels vary because of randomly changing use and loss components
superimposed on the leakage level. A statistical estimate of the leakage level may be obtained by
analysing the variation of these random components.
Socrates
Distribution penetrates
intermittent use
of Socrates
(random) to
minimum
micro-values ‘core’ flow
over
series of
continuous ‘use’
night-lines random
plumbing losses
mostly
WASTE
leakage (unseen holes)
The Socrates minima – the core flow levels – will be distributed about some mean value (m). The
spread of this distribution can be measured by its standard deviation (sigma). We can estimate the
position of the leakage level as being some multiple of sigma down from the mean. In our analysis we
take the leakage position as being 2.5xsigma down from the mean. This is standard statistical
methodology where the sigma multiplier is established as being in the range 2.5 to 3. We use 2.5 as
being the most cautious value in that range. To be extremely conservative, a value of 2 could be used.
a) Outlier Rejection
Sometimes Socrates minima appear abnormally high compared with their peers in the sequence of data
being evaluated. These values are termed ‘outliers.’ A reason for this may be an unusual amount of use
during that night. If we include such outliers in our analysis, the mean level of the distribution rises but
so too does its standard deviation. The net effect is to generally produce a leakage estimate (= m –
2.5xsigma) which is unduly low because of the positive skew introduced by including the outlier. In
producing an estimate of leakage, we need a representative set of minimum values that describe the
usual pattern of events in the area. Unusually high values (outliers) distort that estimate. Positive
outliers may be identified as being those values which have a ‘standardised residual’ between 2 and 3.
(A value’s standardised residual is the value minus the mean and all divided by the standard deviation).
Again, erring on the cautious side, we reject values with outliers with residuals greater than 2.
Negative outliers are also possible. Some minimum values may appear unduly low. Causes include
surge flows which have not been trapped by Socrates’ filters. Low minimum values however may be due
to genuine periods of low use. It is possible to inspect the lowest minimum values using Socrates’
snapshot facility to ascertain whether these values are likely to be genuine or false. False values are
manually removed.
b) Residual skewness
Even after outliers have been removed, it is rare for a distribution of minima to be symmetrical about its
mean; a degree of positive or negative skewness may persist. On some distributions it is possible to
Copyright 2000 Primayer Ltd – All Rights Reserved Page 9
Socrates User Manual Issue 2 Theory of Operation
reduce the effect of skewness by transforming the distribution to a log distribution. To do this logarithms
are taken of each minimum value, and the mean and standard deviation calculated for these transformed
values. The log of the estimate of the position of the leakage for the transformed distribution is the mean
of the log values minus 2.5 times the standard deviation of the log values. The linear estimate (i.e. non
log value) is then taken as the anti-log of the log estimate.
To decide whether to use a linear or log estimate, the system calculates the skewness of both types of
distribution. The distribution with a skewness value closest to zero is chosen.
Both sets of data (minimum flows and two hour flows) are pressure adjusted about the mean nightline
operating pressure of the site. The pressure adjusted values are then used in the calculation described
in 2.6.6 above.
A more ‘cautious’ alternative approach exists to the statistical method described above. This is termed
the Effective Minimum method. The effective minimum is determined from the Socrates nightly minima
(over the sampled period) but without using the Socrates statistical analysis, by applying the following
logic.
If a lower nightly minimum is recorded, after previously higher values, then this lower minimum can be
extrapolated back in time because flows through leaks do not reduce over time. Therefore, higher
figures must be due to use.
This logic can be applied as long as no leaks have been repaired during the period of analysis. Using the
effective minimum (for an individual night) the legitimate night use is calculated (by subtraction of effective
minimum flow from average two hour flow value. An average is then taken over the sample period.
Detailed four hour nightline flow measurements were taken over two week periods from six consumption
monitor sites. Using data from each site, large DMA’s were simulated by adding together randomly
selected two hour nightline segments. Because the core flow level was visible in each data segment, the
effect on combined core level visibility could be characterised as segments were merged together.
For each site, 10 random simulations were carried out testing 5 different averaging times per simulation
for 8 property counts (500 to 4000). The results are shown in Table 1.
Micro-period (seconds)
15 30 60 90
120
Property Count
500 0.016 0.017 0.020 0.023 0.025
1000 0.047 0.051 0.061 0.073 0.085
1500 0.098 0.107 0.129 0.146 0.164
2000 0.150 0.165 0.193 0.216 0.242
2500 0.217 0.236 0.269 0.300 0.327
3000 0.293 0.314 0.350 0.392 0.425
3500 0.356 0.384 0.427 0.474 0.517
4000 0.422 0.458 0.507 0.569 0.605
Table 1 Average differences between actual minima and Socrates minima for different
property counts and averaging times (in litres/second).
A practical example illustrates the use of this table. Consider a Socrates minimum measurement taken
on a 1000 property DMA using an micro-period of 30 seconds. The table suggests that the measured
value will be too high by 0.051L/sec. This error is called the penetration error.
To build in ENU’s, we need measurements or estimates of nightly continuous minimum flows into each
of them over the analysis period. These can be obtained from the following sources:
It must be realised that measurements obtained at options ii) and iii) will include inaccuracies, but if the
ENU flows are much lower than the minimum flow measured by Socrates into the DMA, then the overall
error is minimised. Each ENU’s nightly minimum value, together with overall site penetration error, is
subtracted from the corresponding Socrates minimum measurement for the DMA. The resulting values
are termed the Net Minima. When any outliers have been rejected (see 2.6.5 above) the resultant
values are termed the Qualified Minima and these are then analysed using the statistical process
described at Section 2.6.4 above to derive leakage and legitimate night use estimates.
3 Installation
Socrates is housed in a ruggedised cast aluminium enclosure which is fully submersible and is sealed to
IP68. Internal sealing allows the batteries to be changed whilst maintaining the integrity of the system.
3.1 Connections
Kent Helix
and PSM,
Socam,
Meinecke Existing
Meters Instrumentation
Pulses in
ABB Kent- Pulses out
4 - 20 mA
Taylor
Magmaster RS232
Meter
Socrates
External
Pressure
Transducer
Connect the pulse unit to the flowmeter. If pulse units are to be stacked then the pulse unit connected to
Socrates must be next to the flowmeter i.e. in the lowest position. A “pass-through” port is provided to
support existing pulse measuring equipment without the complication of stacking pulse heads.
Pressure measurement is by external transducer. The transducer should be selected to provide the
highest possible measurement resolution; e.g. if measuring the pressure in a main supplied by a
reservoir with a head of between 10 and 12 metres, the transducer range should be 2 bar. If it is
necessary to connect the pressure transducer where another transducer is already installed, a T-piece is
available from Primayer to enable both transducers to be connected to a single pressure port.
The communication port operates at 19600 baud to transfer data rapidly (approximately 6 minutes for 1
month of data).
Sensors supplied by Primayer are factory-fitted with sealed matching connectors. Connectors fit with a
bayonet (twist-and-click) action of the rotating shell. All connectors are polarised such that it is
impossible to connect the wrong input, with the exception of the pressure and analogue inputs which
(although not recommended) can be interchanged.
3.2 Sealing
The connectors fitted to Socrates are waterproof only when mated with a matching connector, or when
the captive sealing cap is fitted. Connectors will only seal correctly if dirt is excluded from the mating
faces. It is strongly recommended when fitting sensors that the sealing caps attached to the connectors
are also mated to keep out contamination (in other words, first mate the sensor connector with the
Socrates connector, then connect the sealing caps together).
Socrates sensors operate at low levels of current, consequently any penetration of moisture is likely to
cause electrical leakage causing degradation or destruction of data. Contaminated connectors may be
cleaned by scrubbing with a brush using industrial alcohol (methylated spirits).
Analogue
flow input
Pressure
transducer
input
Pulse flow
input
Pulse pass-
through output
Communications
port
3.3 Use
Socrates is set-up by running the PC host software. Refer to Section 4 of this manual. No routine action
is required when Socrates is in use. Data may be downloaded at any time without interrupting the
recording of data.
To check that Socrates has been installed correctly, a portable computer may be connected to the
Communications Port to allow the host software to display the instantaneous values of flow and pressure
in the Status Window (see Section 4.2.2). A status LED is provided. This flashes when the instrument
has been programmed and is ready for use or is operating.
NOTE: If the Socrates unit is programmed on-site and an Autocal of the pressure transducer is
performed, it is essential that this is done prior to connecting the transducer to the pressure source (i.e.
the Autocal must be done with the transducer connected to atmosphere.)
It is highly recommended that Socrates be installed for a minimum period of one month.
However, when installed on a new site readback of results after a few days often provides
valuable information.
The batteries should be changed when the warning message appears on the PC host software. This will
occur when there is approximately one month battery life remaining.
A special allen-key tool is supplied with the replacement battery pack. Use this tool to remove the eight
allen screws and separate the two halves of the casting. Disconnect and remove the battery pack. Place
the new battery pack in position and connect to the instrument. Place the battery connector where it will
not be trapped in the seal. Do not tamper with any other part of Socrates.
Check the cleanliness of the seal and re-assemble. Insert all screws and turn until hand-tight. Tighten
each screw a further quarter-turn, in the order 1, 8, 2, 7, 4, 5, 6, 3 as shown below. Do not over-
tighten.
4 PC Host Software
4.1 Installation
The PC host software is supplied on two diskettes.
Insert diskette #1 in the drive and run the installation program Setup.exe on that diskette. (Select Run
from the Start menu and type “A:\Setup.exe”, then click on OK.)
Select the directory for the installation when prompted (the default is usually satisfactory), then change
diskettes when prompted during the installation.
Once installed, the software is run from the Program menu. If desired, an icon for the Socrates software
may be placed on the desktop by the standard Windows 95/98 or NT method.
4.2 Start-up
The View options on the main menu control what is displayed on-screen. The following description
assumes that Toolbar, Status Bar and Status Window are all checked. Any or all of these functions
may be disabled if required, by clicking on the name to remove the check mark.
The next stage is to configure Socrates for operation. This is a procedure involving many parameters;
the details are discussed in Section 4.4
Once Socrates has been programmed and is in use acquiring data, it will be necessary to retrieve these
data from time to time. This procedure is discussed in Section 4.5
Data once recovered may be displayed and printed as graphs and as tabular reports. The procedures for
generating these are discussed in Sections 4.6 and 4.7
Flow Value
Pressure Value
Load from
disk
Print
Report
wizard
The toolbar is “dockable”, namely it may be dragged anywhere on the screen, and may be re-
dimensioned as required. (For example, it may be changed to a vertical layout on the right side of the
screen.)
4.3 Tools
4.3.1. Options
The Tools - Options menu selection allows the default settings to be changed. It is important that the
COM port setting is correct, as otherwise the software cannot communicate with Socrates.
a) The General tab allows the database path, the global flow unit selection, the default modem
type (if installed) and the default COM port to be set. Note that the COM port is that to be used
for direct connection to Socrates; this may be the same port as is used for an external modem,
but cannot be the port dedicated to a modem fitted internally to the computer.
Data retrieved from Socrates is saved by default in \Primayer Database \Socrates, and it is
recommended that this is not changed. However, a different directory may be used for the
database if desired; either enter the name of this directory, or select it using the Browse button,
which displays the contents of directories.
The Reported Units selection sets the unit of flow which is used throughout the software i.e. all
graphs and reports use this unit of measurement.
b) The Graph tab allows the trace colours, the default graph duration, the graph title and the font to
be selected. To select a colour for a trace or for the graph background, click the desired colour
square and drag it to the trace or background.
c) The Report tab allows the report title, the default report duration and the report font to be selected.
It also allows for specific settings relating to the ‘Interpretation of Results’ report, as follows:
i) Normally Socrates will automatically carry out an evaluation of whether a logarithmic or linear
estimate gives the best removal of any ‘skewness’ in the core flow distribution (see Theory of
Operation 2.6.5(b)). The use of logarithmic values in the removal of ‘skewness’ can be
disabled. If disabled then only a linear estimate will be carried out. It is recommended that
this box remains in the ‘ticked’ condition.
ii) The standard deviation multiplier used to determine leakage from the core flow distribution
(see Theory of Operation 2.6.4) defaults to a value of 2.5. However, this value can be
adjusted if required.
iii) The standardised residual value is used for eliminating positive outlier values before
determining leakage from the core flow distribution (see Theory of Operation 2.6.5(a)). This
defaults to a value of 2.0. It can be adjusted if necessary.
The names of the parameters and their effects are discussed in Section 2 – Theory of Operation. It is
recommended that no changes are made to the Transient Detection parameters unless the underlying
theory is understood.
The various parameters for a Socrates setup are defined by stepping through a series of screens, which
are accessed by the tabs at the top of the window. These parameters can then be saved for later recall,
or immediately programmed into Socrates, or both. Alternatively, a setup can be retrieved from the
database or from Socrates itself, modified if necessary and then used as desired.
Set-ups are stored in the Primayer database which is automatically created when the software is
installed. The Setup Name is the key used to index these set-ups. (When data is retrieved from
Socrates, the sets of data are indexed by the Site name.)
The buttons at the foot of the window are active when any tab is displayed. The OK button saves the
setup; the Program Now button allows data to be transmitted to the Socrates analyser. It is, of course,
essential that all parameters on all tabs are verified before programming Socrates
a) The Setup tab enables new set-ups to be recorded, previously created set-ups to be retrieved,
old set-ups to be deleted, or an existing setup to be read from a Socrates analyser. The
Comms Setup button allows the data handling port to be defined.
b) The Flow Meter tab selects the flow meter to be used with the Socrates analyser. The
parameters displayed change according to the type of meter selected.
When an inferential meter, such as a Kent Helix type or Socam/Meinecke WS or WP meter,
is selected then typically the following information must be entered for operation of Socrates:
Meter size <= 100mm
>= 150mm
Average window flow < 10 litres/sec
} For meter <= 100mm
>= 10 litres/sec
<= 100 litres/sec
} For meter >= 150mm
> 100 litres/sec
Pulse head type HRP/PU100 (preferred) or
LRP/PU10
PD100
FCAL value Enter calibration factor in units of litres/pulse
(For the Socam meters enter only the FCAL value - see Appendix A for all FCAL values.)
The parameters for a PSM-type meter are similar, but there is one important difference. The
Desired accuracy of zero-flow measurement defaults to 2%. However, if very low flows are
anticipated, the value of 5% should be selected.
If an electromagnetic meter type is selected then the following information must be entered:
Note: This setup also applies to the Aquamag meter only if it is powered continuously.
The Histogram Setup button on the Flow meter tab allows the limits for the flow histogram to be
set. The flow histogram shows the distribution of the occurrence of flow values overnight and is
useful to understand when trade or exceptional use has occurred. To obtain the best scaling on
the Socrates histogram it is recommended that the maximum flow value is set to the expected
maximum flow rate during the window period.
c) The Time tab allows the start date and the duration period of the analysis to be set. The analysis
window for each night is set by its start time and duration.
d) The Transient Detection tab sets the Socrates analyser parameters to allow transient effects
to be smoothed. The Average Head of Pressure is selected according to the location of the
Socrates analyser.
The Advanced button allows the transient detection parameters to be adjusted. Note that these
have been found to be satisfactory. The parameters should be changed only where necessary.
An explanation of each parameter is given for guidance.
Qualifying Time for both flow and pressure transients sets the time (in seconds) of a
fixed length window moving in time. A transient is detected if the difference between any
two samples in this window exceeds the threshold set below.
Flow Change Threshold sets the percentage difference between any two flow samples
(in the window defined above) which qualifies as a flow transient. If triggered the data in
the current window is ignored.
Pressure Transient Detector Trip Level sets the difference (in metres) between any two
pressure samples in the window defined above which qualifies as a pressure transient or
surge. If triggered the data is ignored for the period defined below (to allow the surge to
decay).
Pressure Transient Decay Time sets the period (in seconds) for which data is ignored
following a detected pressure transient.
d) The Site Details tab allocates an ID and name to the site. Site ID and Name MUST be entered,
as these are the parameters used to identify the data. The number of properties is used to
calculate the leakage value and legitimate night use value.
e) The Pressure Transducer tab accesses the calibration parameters of the external pressure
transducer. The pressure transducer range should be selected from the list. The PCAL value is
the “Full scale range output” figure on the manufacturer’s calibration certificate. If the transducer
offset is known, this should be entered; alternatively, tick the Perform Autocal box to cause
Socrates to measure the value automatically. The Perform Autocal checkbox runs the Autocal
function (as described at 4.4.1 below) to calibrate the transducer when Socrates is programmed.
It is ESSENTIAL that the transducer is connected and sensing atmospheric pressure when the
instrument is programmed, if Autocal is selected to be performed.
4.4.1 Autocal
Autocal is a method of correcting the drift with time of a transducer, and involves measuring the
transducer output when it is sensing atmospheric pressure (i.e. measuring the zero offset).
This may ONLY be done if the transducer is not sensing water pressure at the time when the Autocal is
performed.
To perform Autocal, first ensure that a transducer is connected to Socrates and that the transducer is
sensing atmospheric pressure, then click on Autocal in the Analyser menu.
When Program Now is selected one final detail is required. If the Socrates is likely to experience
extremely low flows (which it might interpret as a meter ‘stall’ situation) then it needs to be programmed
accordingly. From the flowmeter set-up instructions Socrates will calculate a critical low flow rate (in
litres/second) and this value is displayed. If there is a possibility that the flow may fall below this
extremely low critical value, during the night window analysis period, then press yes so that these low
flow values are recorded. If yes is not selected then the message ‘zero flow,’ will be recorded.
The default data setting is All Data. To select specific data, first un-check the All Data checkbox then
check the desired data. As Socrates retains a record of when data were last read, it is possible to
retrieve only the data which have been stored since that time. To do this, click on Read Data since last
read.
Data recovery can take a considerable time, because of the volume of information. As a guide, reading
30 days of data takes approximately 6 minutes.
When data has been recovered, the option is given to display a graph or report of the data. The graph
option will display a ‘daily graph’ and the report option will display an ‘Interpretation of Results’ report.
4.6 Graphing
4.6.1 Graph Wizard
The Graph Wizard is the function which displays recovered data in graphical formats. To obtain a graph,
follow the steps as shown by the Wizard.
Daily Data plots selected Socrates daily values against time. Several data sets from a single
site may be displayed together as lines of different colours.
Histogram Data displays a one-day histogram of flow data sets. The flow histogram shows the
distribution of the occurrence of flow values overnight and is useful to understand when trade or
exceptional use has occurred.
SnapShot Data displays flow and pressure data, for the 7 nights with the lowest minimum
flows. The flow is displayed for 30 seconds prior to the point of (Socrates measured) minimum
flow and approximately 4 minutes after the point of minimum flow.
Micro-Value plot displays the micro-values over the most recent night window period. Note that
this data is only stored for one day, as it is overwritten every morning (regardless of whether or
not the Socrates analyser is connected to sensors). Therefore to retrieve this data plot it is
necessary to read Socrates before the next window analysis period.
The Site name is displayed, together with the data start and stop dates, for all data present in the
database. Click on the desired name to highlight it, then set the desired start date and duration for
displayed data. Check the boxes on the right to select the desired data.
The type of data to be displayed must be selected. There are two categories - Daily Window Data and
Continuous Logged data.
1) Socrates minimum night flow qualified for flow and pressure transients is the minimum flow
determined after data containing both pressure and flow transient effects have been removed. This
data is determined using measurement and hydraulic artefact correction.
2) Socrates minimum flow qualified for pressure transients only is the minimum flow determined after
data containing pressure transient effects have been removed. This data is similar to the data in 1)
above, but without the removal of flow transient effects. This data is determined using measurement
and hydraulic artefact correction.
3) Minimum night flow (15 minute average) is the minimum daily 15 minute average flow value as would
be determined by counting flow pulses over 15 minutes. It does not have the measurement and
hydraulic artefacts removed. This value is provided for comparison purposes only.
4) Minimum night flow (2 hour average) is the average flow measured during the window, i.e. it is the
average of eight 15 minute flow averages, and so does not have the measurement and hydraulic
artefact correction.
5) Average pressure (2-hour average) is the average pressure measured during the window.
6) Standard deviation is the measure of the spread of the micro-values measured during the window.
7) Number of pressure transients is the number of times data was discarded because the pressure
exceeded the pressure transient detection trip level.
Step 3 titles the graph and allows the axis titles to be changed:
The procedure described above is for Daily Data. The procedure for Histogram Data, Snapshot Data and
Micro-Value Plot is similar, with the following changes:
a) Histogram Data
The selection of data in Step 2 is limited to the Site name and the day for which the
histogram is to be displayed. Only the title of the histogram may be entered at Step 3
b) Snapshot Data
The selection of data in Step 2 is restricted to the Site name and the day for which the graph
is to be displayed. Only the title of the graph may be entered at Step 3. The dotted vertical line
indicates the position of Socrates measured minimum flow.
Micro-value plot
Zoom Time (axis): when selected, clicking the mouse on the graph and dragging left or right produces a
rectangular selection box (extending the full height of the graph) which defines the section of graph to be
zoomed.
Zoom Left (Y-axis): when selected, clicking the mouse on the graph and dragging up or down produces
a rectangular selection box (extending the full width of the graph) which defines the section of graph to be
zoomed.
Zoom Right (Y-axis): operates in the same manner as the left Y-axis zoom.
Zoom All: allows both time and Y-axes to be zoomed together. The selection box is fully defined by the
mouse movement, which selects both height and width of the area to be zoomed.
Redraw
Zoom All
Pan Right
Undo the
last action Query Point
Query Point: measures a point value when selected. The display opened is dockable. Clicking the
mouse on any point on the graph selects that time, and displays the time and the trace value at that
time. If more than one trace is displayed, the up and down arrows in the display scroll through the
values of the different traces, which are displayed in the colours of the traces.
Pan Left/Pan Right: if there is more data than displayed on screen, move the start to left or right. The
change in position is one half of the displayed time - i.e. if the display shows two days of data, a pan will
change the start time of the display by one day.
Text Entry: allows any text to be typed over the graph. The mouse must be clicked at the desired start
position before typing.
Move Text: when selected, the text is surrounded by a box, which is dragged to re-position the text.
Report: displays the on-screen graph in the form of a tabular data report, i.e. the exact data displayed
as a graph is displayed as a data report.
4.7 Reporting
4.7.1 Report Wizard
The report wizard is the function which displays Socrates data in tabular formats. The first screen allows
the report format to be selected:
Interpretation of Results displays daily minimum night flow, together with net and qualified
minimum’s which take account of penetration error and exceptional night users. Over a user-
defined period the leakage estimate and mean legitimate night use are displayed.
Diagnostic Report displays selected Socrates daily values against time. Several data sets,
from a single site, may be selected for printing together.
Logged Data Report displays continuous flow and pressure data logged at 15-minute intervals.
No selection of data types is possible.
SnapShot Data Report displays flow and pressure data for 30 seconds prior to the point of
minimum flow and 4 minutes after the point of Socrates measured minimum flow.
The Site name is displayed, together with the data start and stop dates, for all data present in the
database. Click on the desired name, then set the desired start date and duration for displayed data.
At step 3 the user is provided with the opportunity to enter details of Exceptional Night Users (ENU’s)
accounting for continuous trade use:
The principle of ENU’s was explained in Theory of Operation 2.7.2). Firstly, the number of ENU’s must
be entered. Then the details are entered by three options:
i) Socrates minimum flow measurements at an ENU meter. Here the Socrates data file must be
selected (located from the same database path as defined at the Tools-Options menu).
ii) Estimates obtained from data loggers set at fast sampling (taking care to avoid errors caused by
transients, cyclic oscillation and quantisation). These values are entered manually.
iii) If neither of the above two options is practical, minimum fifteen minute average values may be
used as a last resort. Fifteen minute averaging, however, will exaggerate the extent of trade use.
Again these values are entered manually following the instructions on-screen.
Step 4 allows the user the following further choices in the selection of final details.
ii) An option is provided allow the user to correct for pressure variation to the Net Minimum Flow
values and Qualified Minimum Flow values used in the calculation of leakage and night-use. This
option defaults to not using the pressure correction.
iii) An option to use the Effective Minimum method (see section 2.6.8 for description) as an
alternative to the Socrates statistical method.
Follow the on-screen instructions to display the report. A sample Interpretation of Results report is
shown below. This sample uses the Socrates statistical method, contains two Exceptional Night Users
and is applying pressure correction to the Net and Qualified Minima.
a) The minimum values (over the period selected for display) are shown in bold italics.
b) Where a value fails to meet the selection criteria for the Qualified Minima then no value is
displayed in this column and it is not included in the leakage calculation.
c) Further values can be de-selected manually from the Qualified Minima column (and thus not
included in the leakage calculation) by double ‘clicking’ on the value using the left mouse
button. Alternatively, if a value has been previously de-selected, then double ‘clicking’ will re-
enter the value into the Qualified Minima column and into the leakage calculation.
d) The Leakage Estimate and the Mean Legitimate Night Use (over the selected period) are
shown at the bottom of the report.
e) An indication is provided of the pressure - average flow relationship and the pressure - minimum
flow relationship at the Socrates location. This information is only displayed if ‘pressure
correction’ was ticked at Step 4 of set-up for the Interpretation of Results Report.
f) When the ‘graph’ button is selected (from Report only) the estimate of leakage is displayed on
the resultant graph.
g) Two additional Tool-Bar buttons exist on the Interpretation of Results report (which are not
available on other reports). The first of these allows the dates of the displayed data to be
adjusted. The second allows the ENU entries to be edited.
h) Special messages can be displayed in the Socrates Minimum Night Flow column, as follows:
Zero Flow: This shows that the flow did fall to zero during the analysis window
period and Socrates was therefore unable to analyse data for this night.
Reverse Flow: This shows that the flow did reverse direction during the
analysis window period and Socrates was therefore unable to analyse data for this night.
Overflow: This means that Socrates experienced a very high flow during
the analysis period and was therefore unable to analyse data for this night. The most
likely cause of this is that the meter calibration instructions were incorrect.
No Flow: This indicates that Socrates has not recorded flow during the
period selected for the report. Possible causes of this is that the set-up instructions
were incorrect or that Socrates has not yet reached the date at which recording is due
to commence.
The Site name is displayed, together with the data start and stop dates, for all data present in the
database. Click on the desired name to highlight it, then set the desired start date and duration for
displayed data.
The Daily Window Data selections are displayed as shown above. Tick the boxes on the right to
select the desired data. These are all daily data (i.e. one value per day) determined during the window
period defined by duration start time and duration.
1) Socrates minimum night flow qualified for flow and pressure transients is the minimum flow
determined after data containing both pressure and flow transient effects have been removed. This
data is determined using measurement and hydraulic artefact correction.
2) Socrates minimum night flow qualified for pressure transients only is the minimum flow determined
after data containing pressure transient effects have been removed. This data is similar to the data
in 1) above, but without the removal of flow transient effects. This data is determined using
measurement and hydraulic artefact correction.
3) Minimum night flow (15 minute average) is the minimum daily 15 minute average flow value as would
be determined by counting flow pulses over 15 minutes. It does not have the measurement and
hydraulic artefacts removed. This value is provided for comparison purposes only.
4) Minimum night flow (2 hour average) is the average flow measured during the window, i.e. it is the
average of eight 15 minute flow averages, and so does not have the measurement and hydraulic
artefact correction.
5) Average pressure (2-hour average) is the average pressure measured during the window.
6) Standard deviation is the measure of the spread of the micro-values measured during the window.
7) Number of pressure transients is the number of times data was discarded because the pressure
exceeded the pressure transient detection trip level.
Increase size
Scroll left of font
one page
Display graph of
data
Go to first
page Go to last page
a) Display graph button displays a graph of the data which is shown on-screen in the
report.
b) To change the amount of data displayed in the window, change the font size.
The sizes of fonts used when printing a report are normally set automatically, but if the font sizes are
changed by using the toolbar buttons then these font sizes are used when printing. This controls the
number of values which are printed on a page.
The file is text. It consists of the Socrates setup parameters identified by headings, followed by a side-
by-side listing of the two channels of recorded data. All data is exported, e.g. daily analysis values, 15
minute logged values, flow histogram data, etc.
To select the site(s), highlight each site name in the right window and click on the double-left arrow (<<)
button. If a name is selected in error, it may be removed by highlighting it in the left window and clicking
on the double-right arrow (>>) button.
The data start date and duration are selected from this screen. The exported file is normally stored in the
database directory, as shown, but this may be changed if desired. Click Finish to generate the export
file(s). One file will be generated for each day exported.
Cosmos II 50 -125 1
150 - 300 10
400 - 500 100
Cosmos CD 50 - 100 1
(single jet meter)
Appendix B Airfreighting
Socrates is powered by a battery pack of lithium cells and as such airfreighting is governed by the
appropriate regulations issued by ICAO and IATA.
Socrates with full battery charge or with some charge remaining in the battery (i.e. Socrates still
operates) are subject to airfreighting restrictions. If Socrates is to be transported by air it is strongly
recommended that advice is sort from Primayer Limited, or from a freighting agent, prior to shipment.
Batteries which are discharged or in a condition close to discharge must be removed (and are subject to
correct disposal) prior to shipping.
NOTE: The above only applies to batteries contained within Socrates, not to separate batteries for which
other regulations apply.