pyraven 0.1.6

Interface with the Rainforest Automation RAVEn USB stick

Some code to talk to the Rainforest Automation “Radio Adapter for Viewing Energy” (RAVEn).

Installation

The easiest way would be pip install pyraven

You could also install direct from this repository:

1. Clone this repo

2. In the top-level directory (where this README.rst is) run pip install --upgrade .

Setup

Prerequisites

You should have already bound (paired) your USB stick to your smart meter.

Binding/pairing your smart meter

Unlike many other devices, the RAVEn doesn’t require a specific “bind to the meter” command to be issued to it. Instead, your energy wholesaler or retailer will have to configure your smart meter to bind with the RAVEn stick. Once the meter has been told to bind to the device, there will usually be a 10-minute window during which you simply have to have the USB stick plugged in and close enough to the meter. The LED will stop flashing, and will turn on solidly once the stick is bound.

Different retailers and wholesalers will have different methods of setting up a binding. Some I know of are:

• Powercor/Citipower (Melbourne):

  • They have a “My Energy” portal that you can sign up for at https://customermeterdata.portal.powercor.com.au/customermeterdata/ (correct as of Jan 2017). Once you’ve signed up, there’s a pulldown from your name that allows you to “Manage my HAN devices”, wherein you can bind your RAVEn to the meter.

• Jemena (Melbourne):

  • Use their portal at https://electricityoutlook.jemena.com.au/ to bind your USB device

• AusNet Services:

  • Use the portal at https://connect.ausnetservices.com.au/webui/ to bind your device

• United Energy:

  • they have a startlingly similar portal to Jemena, located at https://energyeasy.ue.com.au/ that you can use to set up a binding

Usage

There’s a simple CLI, used to assist in the development of the library itself. Once your USB stick is set up and bound to your meter, you can use the CLI to monitor the current instantaneous demand reported by your meter:

localhost$ raven
{'status': 'Connected', 'shortaddr': '0xd131', 'link_strength': 100, 'extpanid': 1234567890123456L, 'is_connected': True, 'channel': 11, 'description': 'Successfully Joined'}
{'divisor': 1000, 'summation_delivered': 2220.575, 'raw_summation_received': 0, 'timstamp': '2018-03-27T02:45:45Z', 'raw_summation_delivered': 2220575, 'multiplier': 1, 'summation_received': 0.0}
{'divisor': 1000, 'summation_delivered': 2220.575, 'raw_summation_received': 0, 'timstamp': '2018-03-27T02:45:45Z', 'raw_summation_delivered': 2220575, 'multiplier': 1, 'summation_received': 0.0}
{'timestamp': '2018-03-27T02:47:56Z', 'raw_demand': 142, 'multiplier': 1, 'divisor': 1000, 'demand': 0.142}
...

and so on.

Instantaneous readings show the current demand, and the summation includes power delivered to the grid (eg from a PV array).

The frequency of different types of data delivery are set by the schedule in the USB stick. By default, they are:

• Instantaneous demand: 8 seconds

• Summation: 240 seconds

• Profile data: disabled

• Scheduled prices: 90 seconds

• Price: 90 seconds

• Messages: 120 seconds

• Time: 900 seconds (reports the current time known to the meter)

However, the meter I have has no pricing data (this seems to be common in Australia, because wholesalers and retailers are different entities, and the retailer applies the pricing based on their own peculiar calculations from the raw consumption data) and so the price elements are never emitted.

Similarly, if there are no messages to be consumed, the RAVEn will not emit a message element.

In practice, this means that you’ll usually only get the instantaneous demand and summation outputs.

Documentation

The XML protocol specification can be downloaded from Rainforest Automation

Other works

There are other open-source projects which cover some of the same ground. My intents in re-inventing this wheel were to provide a more complete and standardised Python implementation that could be used as a library by others, instead of the fairly implementation-specific versions already out there.

• Entropy

  • Python

  • streams data to plot.ly

• python-raven

  • Python

  • publishes data to a Mosquitto server (MQTT)

• node-raven

  • NodeJS

  • publishes to an MQTT server