home.newts.org is a tiny  Raspberry Pi Linux  computer.  Currently it runs pyweather to connect my Davis Vantage Pro2 weather station to wunderground.com.  It also runs my tf mud client in a screen session and has a Cherokee web server that consolidates some information about the house and the solar power generation.

http://home.newts.org:8088/cgi-bin/home.py

The page uses python to get XML data from the Vera Z-wave controller and thermostat and HTML data from the Enphase Envoy gateway controller.

#!/usr/bin/python
 
import xml.etree.ElementTree as ET
import urllib2
import re
from bs4 import BeautifulSoup
import time
 
page = open('index.html').read()
 
now = time.localtime(time.time())
day = now.tm_mday
year = now.tm_year
month = now.tm_mon
 
# get the inside temperature
#
vera = 'http://192.168.1.4:3480/data_request?id=lu_sdata&output_format=xml'
vera_xml_data = urllib2.urlopen(vera).read()
root = ET.fromstring(vera_xml_data)
 
units = root.findall(".")[0].attrib['temperature']
thermostat = root.findall(".//devices/device[@name='Thermostat']")[0]
temperature = thermostat.attrib['temperature']
setting = thermostat.attrib['heatsp']
inside_temperature = 'Temperature = %s°%s Thermostat = %s°%s' % (temperature,units,setting,units)
 
# Get the solar production data
#
solar_home_page =  urllib2.urlopen('http://192.168.1.232/home').read()
solar_production_page = urllib2.urlopen('http://192.168.1.232/production?locale=en').read()
 
solar_re = re.compile(r'<!-- START MAIN PAGE CONTENT -->.*<!-- END MAIN PAGE CONTENT -->', re.MULTILINE+re.DOTALL)
 
solar_home = solar_re.findall(solar_home_page)[0]
solar_production = solar_re.findall(solar_production_page)[0]
 
soup = BeautifulSoup(solar_home)
tables = soup.find_all('table')
solar_statistics = tables[1]
 
solar_data = """
%s<br>
%s<br>
""" % (solar_statistics,solar_production)
 
# this was an attempt to display the nice graphs from wunderground.com
# it works if you go to wunderground.com first to trigger the file to be
# created but won't work otherwise.
#
graph_data = '<img src="http://www.wunderground.com/cgi-bin/wxStationGraphAll?day=%d&year=%d&month=%d&ID=KNHNASHU20&type=3&width=500&showtemp=1&showpressure=1&showwind=1&showwinddir=1&showrain=1">' % (day,year,month)
 
graph_data = '<h2>%s</h2>\n%s' % (time.asctime(now),graph_data)
# grr graph data generation has to be triggered by hitting wunderground
graph_data = ''
 
# Generate the web page from the template
#
print 'Content-type: text/html\n\n'
print page % (graph_data,inside_temperature,solar_data)

I used to believe that I couldn’t really get into solar at my house because of the angles, trees etc.  But now I see they have micro grid-tie inverters (GTIs)  on the order of 100s of watts so that small systems are practical.  In some cases the inverter is even part of the panel.    So I think I can get a few 250 watt panels to offset a little of my energy usage.  The nice thing about these AC panels is that they are stand-alone and able to scale by simply adding more panels.

To compare panels I have a google docs sheet.  In trying to determine how much power I might expect to get out of the panels I made a guess of the amount of sun on average and also a “performance factor” which is real world efficiency based on angles and weather, etc for our latitude and weather patterns.  I made a guess of 8 hours and 60%.  The Watt’s have a great solar array and lots of real time data.  Looking at their yearly data I calculated a “performance factor” of  about 40%.  And in thinking about it more the average daily sun doesn’t matter in my spreadsheet.  It is already factored in without knowing the hours.

This article is just the starting point and has the links to my data sources and calculations.  I still need to research the tax incentives and ponder the different options for panels and inverters.  The built in GTI and racking is nice but expensive.  The external Renogy ones have an optional internet interface (gizmo!).  More on the options later.