Search

Recent Entries

Geographic Information Systems

• OpenSource GIS Review Report

Intelligent Valve Monitor

• IVM6000 - Interpreting results (freezing)
• How does the IVM period relate to alarms?
• Inlet No Outlet error or warning

LBOX

• Configuring TFTP
• FAQ - My laptop doesn't have a serial port, can I still communicate with the LBOX?
• FAQ - What webserver is available for the LBOX?

LX-1 Dragsled

• Setting LX-1 Dragsled for external access.
• New Feature: DART Dragsled DHCP Server
• New DART Dragsled charging indicator

Mesh Box

• Mesh networking in simple terms.

On-site Wastewater Treatment

• Caissie Cape Wastewater Management Study
• Belledune Wastewater Survey
• Caissie Cape Wastewater Study Survey

OSS

• Introducing Scribus
• Scribus Desktop publishing for Linux
• Digital cameras and Kinux

Intelligent Valve Monitor

January 17, 2006

We were recently asked:

> I do have one question: Does the timer reset itself, the reason I ask
> is if the valve was operating normally for say two years and then
> suddenly started to malfunction It would take some time for the
> percents to vary, delaying the alarm, or are the percents calculated
> after each rotation forcing an alarm immediately?

Answer: Certain types of alarm conditions would cause it to alarm more or less "right away". For example if it detects the valve is dosing the same zone/outlet over and over (several times in one day or in a row) or if it detects the valve is skipping one or more outlets every time it would alarm right away.

To capture an "uneven distribution" type of alarm condition, the percentages are calculated over a *sliding* operator set "period". So for a high cycling system you might want the monitor to calculate the average on a daily basis. For certain types of facilities where the wastewater flow rate will vary a lot throughout the week (ie. schools, restaurants) you might want to set the averaging period to 7 days, one month or more. The actual alarm condition is checked at regular intervals within the sliding period.

So a sliding period of 7 days means "the last 7 days" as opposed to "every 7 days". So if the percentages go out of whack within the last seven days it will pick that up.

The IVM will alarm whenever the % variation of the data within that last period goes outside the user-configurable parameter "percent variation". So for some systems a 15% variation might be acceptable but for others a 5% variation might be desired. There will be a balance between avoiding unnecessary alarms for minor "glitches" or variations that are not important and ensuring that if there is a serious problem the operator is notified.

We include some suggestions for how to best set these user configurable parameters in the user manual, or the default settings can be used to start off.

TrackBack URL for this entry: