Monitoring Water Hammer - Revision history

Rhg: /* Very Fast Monitoring */

2022-08-04T23:35:12Z

Very Fast Monitoring

@@ Line 46: / Line 46: @@
 == Very Fast Monitoring ==
 To understand the limits of this particular method of reading values from the DP sensor we then removed fixed times, and instead read values continuously.  This provides readings every 30ms, which is just over 3 times faster still.
 == Software ==

JL: typo

2022-08-02T12:35:44Z

typo

← Older revision		Revision as of 13:35, 2 August 2022
Line 33:		Line 33:
	[[File:Fastdp4.png\|frameless\|693x693px]]		[[File:Fastdp4.png\|frameless\|693x693px]]

	This spike lasts for 4/10th is a second, which in terms of hammer is a long time.		This spike lasts for 4/10th of a second, which in terms of hammer is a long time.

	With the higher resolution monitoring we are also catching more smaller spikes in DP.		With the higher resolution monitoring we are also catching more smaller spikes in DP.

Rhg: /* Fast Monitoring */

2022-08-02T12:32:30Z

Fast Monitoring

← Older revision		Revision as of 13:32, 2 August 2022
Line 28:		Line 28:

	At 1/10th of a second readings we see more detail, including the drop-off in pressure following the hammer. This confirms that we are seeing a genuine pressure spike, and also that the system is been exposed to extremely high differential pressures.		At 1/10th of a second readings we see more detail, including the drop-off in pressure following the hammer. This confirms that we are seeing a genuine pressure spike, and also that the system is been exposed to extremely high differential pressures.

			A second event, a day later...

			[[File:Fastdp4.png\|frameless\|693x693px]]

			This spike lasts for 4/10th is a second, which in terms of hammer is a long time.

	With the higher resolution monitoring we are also catching more smaller spikes in DP.		With the higher resolution monitoring we are also catching more smaller spikes in DP.
Line 37:		Line 43:

	We have identified the reason why valves are letting by. The next step is figuring out the cause of the spike.		We have identified the reason why valves are letting by. The next step is figuring out the cause of the spike.

	~~A second event, a day later...~~

	~~[[File:Fastdp4.png\|frameless\|693x693px]]~~

	~~This spike lasts for 4/10th is a second, which in terms of hammer is a long time.~~

	== Very Fast Monitoring ==		== Very Fast Monitoring ==

Rhg: /* Fast Monitoring */

2022-08-02T12:30:34Z

Fast Monitoring

← Older revision		Revision as of 13:30, 2 August 2022
Line 27:		Line 27:
	[[File:Fastdp2.png]] [[File:Fastdp3.png]]		[[File:Fastdp2.png]] [[File:Fastdp3.png]]

	At 1/10th of a second readings we see more detail, including the drop-off in pressure following the hammer. This confirms that we are seeing a genuine pressure spike, and also that the system is been exposed to extremely high differential pressures. We have identified the reason why valves are letting by. The next step is figuring out the cause of the spike.		At 1/10th of a second readings we see more detail, including the drop-off in pressure following the hammer. This confirms that we are seeing a genuine pressure spike, and also that the system is been exposed to extremely high differential pressures.

			With the higher resolution monitoring we are also catching more smaller spikes in DP.

			[[File:Fastdpspikes.png\|frameless\|1224x1224px]]

			[[File:Fastdpspikes2.png\|frameless\|1212x1212px]]


			We have identified the reason why valves are letting by. The next step is figuring out the cause of the spike.

	A second event, a day later...		A second event, a day later...

Rhg at 11:37, 1 August 2022

2022-08-01T11:37:03Z

← Older revision		Revision as of 12:37, 1 August 2022
Line 16:		Line 16:
	[[File:Dpmon3.png\|frameless\|1056x1056px]]		[[File:Dpmon3.png\|frameless\|1056x1056px]]

			== Fast Monitoring ==
	In response, the logging was increased to 1/10th of a second. This is too rapid for discrete values to be communicated, so instead we select to store values into a shift register of 50 values (5s) and send this out as a complete CSV file when a spike is detected.		In response, the logging was increased to 1/10th of a second. This is too rapid for discrete values to be communicated, so instead we select to store values into a shift register of 50 values (5s) and send this out as a complete CSV file when a spike is detected.

Line 33:		Line 34:

	This spike lasts for 4/10th is a second, which in terms of hammer is a long time.		This spike lasts for 4/10th is a second, which in terms of hammer is a long time.

			== Very Fast Monitoring ==
			To understand the limits of this particular method of reading values from the DP sensor we then removed fixed times, and instead read values continuously. This provides readings every 30ms, which is just over 3 times faster still.

	== Software ==		== Software ==

Rhg at 11:17, 1 August 2022

2022-08-01T11:17:07Z

← Older revision		Revision as of 12:17, 1 August 2022
Line 26:		Line 26:
	[[File:Fastdp2.png]] [[File:Fastdp3.png]]		[[File:Fastdp2.png]] [[File:Fastdp3.png]]

	At 1/10th of a second readings we see more detail, including the drop-off in pressure following the hammer. This confirms that we are seeing a genuine pressure spike, and also that the system is been exposed to extremely high differential pressures.		At 1/10th of a second readings we see more detail, including the drop-off in pressure following the hammer. This confirms that we are seeing a genuine pressure spike, and also that the system is been exposed to extremely high differential pressures. We have identified the reason why valves are letting by. The next step is figuring out the cause of the spike.

	~~We have identified the reason why valves are letting by~~. ~~The next step~~ is ~~figuring out the cause~~ of ~~the spike~~.		A second event, a day later...

			[[File:Fastdp4.png\|frameless\|693x693px]]

			This spike lasts for 4/10th is a second, which in terms of hammer is a long time.

	== Software ==		== Software ==

Rhg at 15:36, 31 July 2022

2022-07-31T15:36:33Z

← Older revision		Revision as of 16:36, 31 July 2022
Line 2:		Line 2:
	Monitoring differential pressure is one of the most critical functions in any heat or plumbing network.		Monitoring differential pressure is one of the most critical functions in any heat or plumbing network.

	DP is required to make water flow, but too much will damage equipment.[[File:Dpgraphtrend.jpg\|thumb\|443x443px\|Existing BMS systems historical data at 15 minute intervals]]One of the major problems with current BMS systems is ~~there~~ speed - they are far too slow to spot spikes in pressure. BMS systems typically report values every 15 minutes, so trying to understand an event that may last only 0.5s is impossible. The top image on the right is from a typical BMS system, where the frequency of data logging is far too slow to catch hammer events. From this data everything looks fine, with DP controlled to targets. But the important problems are hidden - on this site something is forcing valves to let by.		DP is required to make water flow, but too much will damage equipment.[[File:Dpgraphtrend.jpg\|thumb\|443x443px\|Existing BMS systems historical data at 15 minute intervals]]One of the major problems with current BMS systems is their speed - they are far too slow to spot spikes in pressure. BMS systems typically report values every 15 minutes, so trying to understand an event that may last only 0.5s is impossible. The top image on the right is from a typical BMS system, where the frequency of data logging is far too slow to catch hammer events. From this data everything looks fine, with DP controlled to targets. But the important problems are hidden - on this site something is forcing valves to let by.

Rhg at 15:31, 31 July 2022

2022-07-31T15:31:36Z

← Older revision		Revision as of 16:31, 31 July 2022
Line 20:		Line 20:
	We then waited for another event.		We then waited for another event.

			A day later the following email arrived.

			[[File:Fastdp1.png\|frameless\|339x339px]]

			[[File:Fastdp2.png]] [[File:Fastdp3.png]]

			At 1/10th of a second readings we see more detail, including the drop-off in pressure following the hammer. This confirms that we are seeing a genuine pressure spike, and also that the system is been exposed to extremely high differential pressures.

			We have identified the reason why valves are letting by. The next step is figuring out the cause of the spike.

			== Software ==
	The flow below runs on a standard Heatweb open controller, and monitors differential pressures at 0.1s intervals. When pressure out of limits is detected, 50 consecutive readings (5 seconds worth) are emailed to engineers concerned for record and analysis, with the spike positioned in the middle of the data so we can see climb and fall in pressure.		The flow below runs on a standard Heatweb open controller, and monitors differential pressures at 0.1s intervals. When pressure out of limits is detected, 50 consecutive readings (5 seconds worth) are emailed to engineers concerned for record and analysis, with the spike positioned in the middle of the data so we can see climb and fall in pressure.

Rhg at 16:56, 30 July 2022

2022-07-30T16:56:19Z

← Older revision		Revision as of 17:56, 30 July 2022
Line 2:		Line 2:
	Monitoring differential pressure is one of the most critical functions in any heat or plumbing network.		Monitoring differential pressure is one of the most critical functions in any heat or plumbing network.

	DP is required to make water flow, but too much will damage equipment.[[File:Dpgraphtrend.jpg\|thumb\|443x443px\|Existing BMS systems historical data at 15 minute intervals]]One of the major problems with current BMS systems is there speed - they are far too slow to spot spikes in pressure. BMS systems typically report values every 15 minutes, so trying to understand an event that may last only 0.5s is impossible. The image ~~below~~ is from a typical BMS system, where the frequency of data logging is far too slow to catch hammer events. From this data everything looks fine, with DP controlled to targets. But the important problems are hidden - on this site something is forcing valves to let by.		DP is required to make water flow, but too much will damage equipment.[[File:Dpgraphtrend.jpg\|thumb\|443x443px\|Existing BMS systems historical data at 15 minute intervals]]One of the major problems with current BMS systems is there speed - they are far too slow to spot spikes in pressure. BMS systems typically report values every 15 minutes, so trying to understand an event that may last only 0.5s is impossible. The top image on the right is from a typical BMS system, where the frequency of data logging is far too slow to catch hammer events. From this data everything looks fine, with DP controlled to targets. But the important problems are hidden - on this site something is forcing valves to let by.

Rhg at 16:55, 30 July 2022

2022-07-30T16:55:53Z

← Older revision		Revision as of 17:55, 30 July 2022
Line 2:		Line 2:
	Monitoring differential pressure is one of the most critical functions in any heat or plumbing network.		Monitoring differential pressure is one of the most critical functions in any heat or plumbing network.

	DP is required to make water flow, but too much will damage equipment.		DP is required to make water flow, but too much will damage equipment.[[File:Dpgraphtrend.jpg\|thumb\|443x443px\|Existing BMS systems historical data at 15 minute intervals]]One of the major problems with current BMS systems is there speed - they are far too slow to spot spikes in pressure. BMS systems typically report values every 15 minutes, so trying to understand an event that may last only 0.5s is impossible. The image below is from a typical BMS system, where the frequency of data logging is far too slow to catch hammer events. From this data everything looks fine, with DP controlled to targets. But the important problems are hidden - on this site something is forcing valves to let by.

	One of the major problems with current BMS systems is there speed - they are far too slow to spot spikes in pressure. BMS systems typically report values every 15 minutes, so trying to understand an event that may last only 0.5s is impossible. The image below is from a typical BMS system, where the frequency of data logging is far too slow to catch hammer events. From this data everything looks fine, with DP controlled to targets. But the important problems are hidden - on this site something is forcing valves to let by.


	~~[[File:Dpgraphtrend.jpg\|thumb\|443x443px\|Existing BMS systems historical data at 15 minute intervals]]~~
	[[File:Dpmon1.png\|440x440px\|thumb\|Wiring diagram for open control monitoring of DP]]		[[File:Dpmon1.png\|440x440px\|thumb\|Wiring diagram for open control monitoring of DP]]
	[[File:DPI.jpg\|thumb\|Grundfos DPI Differential Pressure Sensor]]		[[File:DPI.jpg\|thumb\|Grundfos DPI Differential Pressure Sensor]]