Doris is a computer simulation designed to explore the use of sustainable energy by a typical household.
Doris lives in a simplistic and imaginary world, it poles weather reports from a randomly selected airfield which is near an imaginary wind farm. it has access to some fictitious local solar panels and thinks it is connected to the national grid. The meteorological data is used to estimate the wind and solar power that might be available. There is some storage built into the system and when more energy is available than is required to meet the load, it used to charge batteries. When the wind does not blow and the sun does not shine, the load is supplied from stored energy and when that is consumed it falls back on conventional sources from the national grid. Energy from the wind farm is supplied from the grid whilst it is assumed that the solar panels are located on the consumer's site or on a local grid.
 |
| Doris - What is being simulated |
It is the result of two activities:
- Having spent much of life in the oil and gas industry, I felt the need to explore sustainable sources. This included the academic study of things like wind speed distributions, climate, sun-earth geometry and the effect of clouds on solar irradiance. There was also a practical element which included a 4.5 watt PV panel, a lead acid battery and an LED light known as the Solar Bucket plus a few cycle rides around the town and surrounding countryside with a simple wind speed meter.
- In 2007, before the financial crisis, rooftop PV panels and wind turbines were becoming available and I became curious to to know if it was possible to economically reduce household reliance on the grid by using these products.
Both of these efforts came to the same conclusion, that increasing the proportion of household energy coming from sustainable sources was a challenge. Some caveats are necessary before making any conclusions:
- Economics are important, any alternative to fossil fuels must deliver the same benefits for a comparable cost, if it does not, no one will adopt it.
- The technology is the key technology because it separates production and consumption into two separate processes.
- Much discussion of sustainability focuses on generation, yet managing and reducing consumption is, maybe, more important. Sustainability is easier to attain if the demand is low.
- The nature of a sustainable solution is dependent on the climate in which it is installed, thus the experience of one country may not be relevant to another. Solar PV might work well in the deserts of Arizona, but less so in the highlands of Scotland where wind turbines are a more attractive option.
The conclusions I came to for a location in the south of England were:
- Solar PV works well in summer, but not in winter. On a sunny summer day, the cumulative GHI might be 8 kwh/m2, whilst on an overcast December day it can be less 1 kwh and that is the time of year is when energy consumption is at a peak.
- The most effective way of generating energy from the wind is from industrial scale wind turbines located in optimum locations such as onshore ridges and offshore.
- Both solar and wind are discontinuous sources of energy, the sun does not shine at night and wind energy comes in pulses which can be several days apart.
- Wind and solar sources will not be able to fully displace fossil fuels in the short term but the proportion from sustainable sources can be increased.
A system which could mitigate some of these issues for a typical household might look like this:
- Focus energy consumption and management on storage with a capacity of 5 - 10 kwh.
- Only draw energy from the grid when equivalent amounts of energy from sustainable sources are being fed into it.
- Ensure that there is access to both wind and solar generating capacity.
- Use energy generated from conventional sources only when the storage is depleted and sustainable source are not available.
Such a scheme might increase the proportion of sustainable energy consumed by a typical house to around 70%.
Doris is software which can be modified to explore ideas without the expense which would be incurred by working with expensive real things. Several configurations and scenarios have been run and these are described in the following posts:
Simulation is not real life, so the results should be treated as a possible direction of travel rather than detailed predictions. At the time of writing, these posts are being edited and are subject to change.
Doris can runs in two modes:
- On a Raspberry Pi with live data from NOAA via an internet connection. The Pi runs the base configuration and is capable of uploading graphics to a shared server. In this case, the behavior of the software has some relation to the weather on the other side of the window.
- Much the same code runs on a laptop using historic data from an SqLite database, this allows different scenarios to be run against the same weather data or similar configurations with different weather data.
The data flow looks like this:
 |
| Doris - Data flow |
Doris is work-in-progress. The first iteration used an old laptop, an interface card and a home brewed bank
of Ni-MH cells, the computer obtained weather data from the internet and made decisions which it implemented using
the interface card and tried to maintain a small load without recourse to the grid (in reality a discarded mobile phone charger). Whilst this was both
entertaining and instructive, it was an inefficient way of exploring the concept, not least because of the energy
overhead of the aging laptop and the risks associated with my dubious knowledge of power electronics.
 |
| Doris - The first attempt, a second would be nicer |
Whilst there are significant technical and commercial issues associated with implementing a "real" version of Doris, it is economics which presents the greatest challenge. Whilst the virtual world of Doris
has the potential to reduce emissions by maximizing the use of energy from sustainable sources, the
economics don't look good. The retail consumer has two pricing options, either a single tariff or
something like Economy 7 which offers electricity for around 7 p/kwh in the wee small hours and at
least twice that during the day. Solar generated electricity is almost by definition can't be bought off peak because the sun does not shine at night. For this scheme to be viable there
needs to be some innovation in the energy market which gives producers are reasonable rate of return
and reduces the reliance on fossil/nuclear sources. One possibility is for users to have a stake in the ownership of generating capacity rather than buying its output at a unit cost.
The base configuration of Doris was chosen such that the installed cost of a non-imaginary system would be within sight of £5k, this is summarized as:
- Annual energy consumption: 2,500 kwh (approx. 7 kwh/day)
- Storage capacity: 10 kwh
- Solar generating capacity: 1 kw
- Wind generating capacity 1 kw
The methodology for estimating solar irradiance under a cloud sky is evolving and is part of a separate project, the method used for Doris has a pragmatic element to it and hopefully will evolve.
Some of the work from which Doris originated is described in these posts:
Who was Doris?
I first heard the expression "I'm Doris, the goddess of wind" from an old bloke I worked with in a sheet metal factory. There's a lot of wit in factories along the lines of "The water is safe to drink because it's been passed by the management". The Doris quip stayed with me and then someone invented the internet and Jimmy Wales started Wikipedia. This tells me that it was the catch phrase of Douglas Byng who is described as an English comic who trod the boards before and after the second world war. As many software titles get subverted, it seemed a good idea to start with something that was already ambiguous.
This post was revised on 03-Nov-2015
