Smart metering for water with the IoT

[Tim Kannegieter]

At 12pm 10 October 2017, this community hosted a webinar will be held on Smart Metering for Water with the IoT. In the comments on this post are some of the questions asked by the audience. Feel free to respond to the questions directly. 

To post a question/comment you need to:

• (register and) logon to this community site in the top right hand corner
• Navigate to Forums > IoT Engineering and locate the post with name of the webinar

 

[Tim Kannegieter]

Can you easily instrument existing mechanical meters? What are the challenges involved?

Answer transcribed from webinar response by Rian Sullings (WaterGroup P/L):

In Australia there are roughly 24 million water meters. Coincidentally, a similar number to the population, so most houses have a couple of people in them, but then if you consider all the other buildings and infrastructure, it adds up to a similar number. The vast majority of those meters are mechanical. They have moving parts. They're similar to a clock. They've got a register (like a car odometer). The meters themselves are designed to last for 10 or 15 years in situ. They wear out over time. They become less accurate.

It is possible to replace an entire water meter with a smart-enabled meter, but it's also possible to retrofit devices on to those mechanical meters to make use of the physical asset that's already sitting there and will likely sit there for years to come. Most of the mechanical meters that have been deployed in Australia for the past decade or two have a provision for a data output. I think the thinking was that, "We don't quite have the technology yet, but we know we will in the future, so let's put data outputs on all the mechanical meters."

The most common way of extracting the data is by attaching a sensor into the meter. If you imagine the register, it's a number of dials and they rotate as the water flows through. On some of those dials there is a magnet and that magnet makes revolutions with the dials or gears. For example, every 10 litres that passes through the meter, a dial might make one full revolution, so then you can use a reed switch or a hall effect sensor to detect when the magnet is close to or further away from the sensor. Then you can count how many times the water meters turns over time. You can use data logging to timestamp that.

[Tim Kannegieter]

You mentioned "new business models for the water industry". What are they?

[PaulD]

Hi Tim and Ryan,

In rural areas are there any other solutions rather than having a repeater every 10km or so?

Thanks

Paul

Answer transcribed from webinar response by Rian Sullings (WaterGroup P/L):

Yes, absolutely. Depending on the geography of the rural area, some of the LPWANs can exceed 10 km reliably, so therefore you'd have a base station that might provide 20 or 30 km if you've got nice flat land. If it's more hilly, then you might have some challenges. Another technology that's starting to emerge and looking like it might be a really viable option for remote smart metering or anything IoT is satellite communications. We've gone from having quite high-powered wireless communication across the land, to the LPWAN or low power overland communication. In the past year or so, there's been a few leaps and bounds made in the low-power satellite communication space where you can, in theory, have one or two D cell batteries on a device, and it can be sending its data up once or twice a day and bouncing it off a satellite and lasting 10, 15 years on those batteries.

 

 

[cjv]

How well does LPWAN handle urgent and important communications? E.g. a water meter detecting a critical failure that needs immediate response.

Answer transcribed from webinar response by Rian Sullings (WaterGroup P/L):

Excellent question. If you compare LPWAN to something we're all familiar with like 3G or 4G, it tends to be quite slow to communicate and send that message, but when I say slow I'm talking in wireless terms. In terms of humans realising information has arrived and reacting to it, it might take a few seconds for that message to travel over the air, and one or two seconds for it to arrive on your PC or on your mobile to give you an alert.

Typically, a smart meter will be set to send data at a regular interval. Depending on the technology, that might be more frequent with smaller amounts of data or it might be less frequent transmissions with larger amounts of data. If all is operating normally, the smart meter will tick along, sending data according to its regular schedule. You can also implement “transmission upon event” where the device has a little bit more cleverness built into it. It might be able to realise the profile of something like a leak, and say: "The water use has being flowing at a continuous rate and it's 3am and it's been doing this for an hour. That's not normal. That could be a leak. I will send a message right now instead of waiting until my next scheduled transmission."

[Steve]

You used the Atamo board for development. Did you use this for the final product? 

Or did you develop your own design? If so, why?

Answer transcribed from webinar response by Rian Sullings (WaterGroup P/L):

We did use the Atamo development board for the prototype, but not for a final product. The Atamo development board is based on Arduino. It has the Atmega microcontroller. What we've ended up using for a final product is the arm.com. They are more power efficient and stable. They do everything that we need them to do. They've got that little bit extra processing power that can future proof things and also enable a bit more of the calculation in the device. Did we develop our own design? We've partnered with electronic designers from other parts of the world.

For us to design a solution in Australia, and have it make business sense, we would have to be distributing it in other parts of the world as well. It makes sense in the short-term for us to do it other way around and partner with those who are in much larger markets than the Australian market, such as the European market, and simply work with them to bring over technology to Australia. However, in doing that, we've had to work with them on improving their designs or adapting them to suit Australia. Obviously, the weather conditions here are significantly different to those in Europe, and in many cases we want more detail of data available than a lot of what's on the European market simply because water is a higher priority in Australia than a lot of European markets. It's not just smart metering for the purpose of getting a meter reading and sending a water bill. It's more required for that network operation, understanding and delivering more detail of data.

[PhilWilkinson]

LORAWAN vs SigFow technology. 

Does anyone have a working preference? 

I have only used LORA.

Answer transcribed from webinar response by Rian Sullings (WaterGroup P/L):

It's a very hotly contested space for the wireless networks. Yes, there's a few companies and organisations out there, as well as a few standards that are out there competing to dominate that space. I get asked at least once a week which wireless network is the best wireless network, and the answer is the one that fits what you're trying to do. It fits your use case and can get you started now, and you'll be able to deliver a project or a solution on time and on budget.

We've used Sigfox and I used that in the example of our prototype. That was simply because it was a lot easier to get started in Australia with Sigfox. There's already an existing network in a lot of the population centres, like the major cities, there's a lot of support online. LoRaWAN, also a lot of support out there and a lot of network coverage, but some of the devices, getting them in Australian frequencies is a little trickier. Not to say it's not possible. It's very possible. You just have to follow it and pursue it and be dedicated to find what you need on the market or build it yourself. And try to avoid the marketing hype that's out there at the moment. Be careful to use impartial sources.

Further information on Low Power WAN page of this wiki.

[Jonathan Anderson]

Is there much value in measuring water parameters other than flow with a single smart meter device, e.g. pressure, temperature, pH?

Answer transcribed from webinar response by Rian Sullings (WaterGroup P/L):

There is value in doing that. Pressure, especially, can be used for gaining insights into network operation. I'm not sure if you'd wire a pressure sensor in every single smart meter across the city, but certainly using smart meters as a vessel to include other sensors to distribute around a network. Maybe every tenth house or so, it might make sense to include pressure to gain operational insights for the network. Also temperature and pH, these sorts of values are very useful in terms of the water quality, making sure that there's no issues with the water as it moves around the pipe network.

Temperature is very commonly included in smart metering devices where the electronics and communications are embedded or integrated into the body of the meter itself, temperature sensors being quite simple and cheap and easy to include. pH, on the other hand, the sensors tend to cost a little bit more and be a little bit more complex. I've not seen yet a pH sensor included in a smart metering device.

Further information on this question can be found in the South East Water case study on this wiki.

[PhilWilkinson]

4 minutes ago, PaulD said:

Hi Tim and Ryan,

In rural areas are there any other solutions rather than having a repeater every 10km or so?

Thanks

Paul

 

 

with LORA at least, that is what you need to do.  However the LORA technology is cheap enough to install transmitters.  Also the coverage in the bush will be further

[PaulD]

If you can get LP satellite comms is there any reason you would use LP wan?

Thanks

Paul

Answer transcribed from webinar response by Rian Sullings (WaterGroup P/L):

Yes. The low power wide area networks over ground may be better in instances where the meter is indoors or doesn't have good visibility to the sky (for example if it is located in abasement car parks or a city like Canberra and Adelaide where there is a metal checker plate over the top of the meter itself). It may also come down to a question of power consumption. The satellite communication for smart metering applications is something that I intend to explore a little bit more, and if there's anybody else out there who's an expert on that or has a cool LP satellite Arduino shield or something like that, feel free to reach out and let me know about it.

 

[Neil Belford]

4 minutes ago, cjv said:

How well does LPWAN handle urgent and important communications? E.g. a water meter detecting a critical failure that needs immediate response.

Possibly the sensor can be configured with business rules to respond on event, rather than in the normal duty cycle

 

[Steve Szyndler]

How is Water Group looking at system upkeep at the physical level (calibration and such)?  Is this left to the client or something being requested as part of ongoing project service?

[WHo]

How do you counter the down time due to vandalism or damage that could possibly give a false alarm (i.e. water leaking)?

[Robin]

3 minutes ago, Jonathan Anderson said:

Is there much value in measuring water parameters other than flow with a single smart meter device, e.g. pressure, temperature, pH?

Any or all of these parameters are useful, particularly for the WQ monitoring. Additional monitoring of the Cl residual is highly desirable.

[Naomi Dickinson]

My question is in relation to the ANU study. Did the $300k saving include the infrastructure maintenance rectification costs, or were those costs captured elsewhere? My query is relation unplanned maintenance costs that may be an initial impediment, despite the long term benefit in timely asset monitoring.

[Clement Ng]

Is water flow,  from mechanical meters, calculated on board the edge device or in platform? 

[PaulD]

8 minutes ago, Neil Belford said:

Possibly the sensor can be configured with business rules to respond on event, rather than in the normal duty cycle

 

But then doesn't that lead to using a lot more power as you need to keep polling for the change in those conditions rather than the scheduled transmission?

[Neil Belford]

2 minutes ago, PaulD said:

But then doesn't that lead to using a lot more power as you need to keep polling for the change in those conditions rather than the scheduled transmission?

Hi Paul - that is possibly true, but my assumption is not to a great extent. I think the significant power use is in making the transmission not doing the reading - in any event I was just minuting as best I could Ryans reply so it was preserved.

[D Cheng]

How do you keep spare smart meters to ensure compatibility ?

[Rian Sullings]

On 10/10/2017 at 1:14 PM, D Cheng said:

How do you keep spare smart meters to ensure compatibility ?

Hi D Cheng,

There are many different meters and many components to check to ensure compatibility. From pipe fittings through to wireless communications and data formats.

Cheers,

Rian

[EvenChu]

Hi...as per my knowledge Water meters are installed at each of the inlets in an apartment, and the totalizers are placed as near as possible. Data from these smart meters is sent to the controller/totalizer, which is then sent to the cloud for further actions. Access to the cloud data via apps enables users to switch on/off the valve. For uninterrupted performance, the meters have a backup power source that can also be solar based. However, the cost of solar based batteries is still quite high and suppliers should keep track of OpEx of the projects to make them viable in the long run.