Contact via email: email@example.com
To those who ask: "How long will it take to pay for itself?"
... Meanwhile there have been 5 years of $0.00 electric bills!!
Located roughly 22 miles ESE of Pittsburgh PA this 18.3 KW Solar Photovoltaic (PV) Array went into service on 1-Dec-2011. Constructed with 80 SolarWorld 255W panels the array provides as much as 20.4 KW of DC capacity feeding into 3 SMA 6000 inverters. This solar plant is expected to produce some 23,000 kWh annually exceeding the energy requirements of the associated residence. We are quite pleased with the results as you can see below. The production numbers come from the hard meter attached to the array which tends to differ very slightly from what is reported electronically by the inverters.
This solar PV system went live on December 1, 2011 and so we consider a year of operation as being from December 1st to November 30th. At the same time we also went live with the geothermal heating and cooling system. A prior year is shown in the table above as a reference. The residence used something like 20,000 kWh a year and around 140 MCF in Natural Gas before the upgrades.
Since in our first year of operation we had not established a credit with the power company we elected to run through at least the beginning of our heating season relying on the gas furnace and not the geothermal heat pump. The latter uses a considerable amount of power and we did not yet know if the solar panels would generate enough to cover its operation. We, of course, relied on that heat pump for cooling the following summer and the heat pump has been employed for all subsequent winters.
We experienced a nice reduction in natural gas usage that can be attributed to the high-efficiency gas furnace and the newly zoned temperature controls. The house is now separated into 7 zones. In prior years there was but one thermostat and we would have to alter duct dampers manually to balance the upstairs and downstairs for heating and cooling seasons.
With the increase in power usage that became apparent during 2013 we determined that the heat pump had been configured for dehumidification. We disabled that function and later decided that normal cooling operations were sufficient in keeping humidity levels under control. The dehumidification is simple to enable when we sense the need. This change did reduce the use of the heat pump and reduced our power usage as we see in the subsequent year's results.
In 2016 we will operate with additional dependence on geothermal heating. By default the system had been configured to switch over to the gas furnace when heating demand could not be satisfied in 15 minutes. We noted that the furnace was being used often. That setting has been changed to 45 minutes. This 5-ton heat pump has the capacity to meet demand on all but the exceptionally cold days. By this change we hope to reduce our Natural Gas usage even further.
This Solar PV system at times generates more power than used by the residence. During these times we push power back on the power company's grid. In effect our meter runs backwards and we build up a credit. At other times, such as through the night, we draw power from the grid and work off the credit. The Grid Status shows the current balance, credit or usage. In general and by design we expect that this solar plant will build up credit.
The system performance and the balance between importing and exporting of power is dependent upon the season. During summer months (June, July, August) the Sun is high in the sky and days are long. On a clear day the system generates considerably more power than is used on a daily basis, even with the geothermal systems busy cooling the residence. On average, power is exported and we augment the kWh credit. During winter months (December, January, February) the days are shortest and the Sun remains low in the southern sky. On rare clear days we may still augment the kWh credit especially if the Sun is strong enough to warm the residence giving the geothermal systems a break from heating. On average however, more energy is required by the residence than is generated and we rely on the accumulated kWh credit which is slowly drained. Given that our operating year begins on December 1st we initially watch the year's energy credit fall negative as we rely on the power company to supply power to cover the generation shortage. Later we rely on the summer to push us into the black and create the excess that we end up exporting for the year.
The performance of the individual solar panels is highly dependent on temperature. The high ambient temperatures of summer coupled with a considerable level of solar heating cuts seriously into the efficiency of the solar panels. During summer months while the available solar energy is tremendous the system cannot achieve peak power performance. The effect of heating is quite dramatic and a reason why the ground mounted panels should be preferred over rooftop mounting systems when space is available. The impact of this can lead to disappointment. Similarly in the winter when panel efficiency is not impacted by heating there unfortunately is limited solar energy. However, even on cloudy days energy is efficiently generated.
For the residence, spring and fall months are times when the demand on the geothermal system for heating or cooling is minimal. Our power needs are correspondingly reduced. In the spring (March, April, May) while the Sun works to bring us out of the cold of winter, cool temperatures and lengthly solar days lead to record system performance. Solar panels are able to reach their peak performance and the majority of the power generated is exported. Our grid status quickly heads out of the red. In fall months (September, October, November) ambient temperatures remain high, the available solar energy has just started to decline, and cool evenings let geothermal systems rest, we again export the majority of generated energy. Our grid credit surges slightly in preparation for the ensuing winter darkness. Warm ambient temperatures however prevent the system from reaching the record peak power levels experienced in the spring.
Our system is not designed to operate off-grid. In fact during power company outages the system is required for the safety of power company workers to shut down. In those events too we are not able to supply enough power to satisfy the neighborhood and so we also shut down to avoid the overload. This occurs automatically.
To run off-grid an energy storage system of some kind is required. This typically is a bank of batteries. For various resons we have decided not to go this way. Instead we use a number of small computer grade UPS systems in the residence to eliminate the annoyance of brief power interruptions (we get many in our area).
Actually, in order for us to run entirely off-grid a serious storage system is required. Based upon our grid status swings, we currently estimate that we would have to be able stockpile the ability to generate some 6,500 kWh in energy. With this in reserve we could completely sever the connection to the power company and operate 24/7 with no more conservation than we already do. This would allow us to run through the winter and then top off the reserves in the summer. There currently is no available technology to do this. It would seem that we could devise a system to store hydrogen generated by fuel cells which would also reuse that hydrogen to generate power. This, however, would require dozens of tanks, present a potential safety hazzard, and would not be appropriate for the residential setting. This technology is not ready for prime time.
There are three paths in which power may flow. The first meter shows the instantaneous power flowing from the Solar PV array on a 20KW scale. The second meter shows the power used by the residence on the same scale. The remaining power, the difference between what is being generated and what we are using, flows onto the grid and that is shown on the third meter. This meter can show both positive and negative power as we may be supplying or using respectively. Note that this meter has a -20 KW to +20 KW scale. Readings on the positive/green side of the scale represent our power company meter running backwards.
Weather directly affects solar production and it can be very localized. You may not be very far from our location and where you are the sun can be shining brightly, nevertheless, our array might be well into the shadow of a cloud. A view of the sky in the background helps to understand the local conditions. Also, the array might be covered with a layer of snow or heavy frost. The live camera image can add reality to the performance data represented by the other displays on this page. Besides, it can provide entertainment as you may catch a passing deer or a jolly homeowner on his lawn tractor or brushing away the snow. The view in this camera is to the ENE of the residence and during summer you can watch the sunrise. Luckly pointing this camera into the morning sun does not damage it.
About the Data
This page displays live data. The table updates every 10 seconds and displays the instantaneous AC power generated, the total accumulated for the current day, the total generated since installation and the balance with the grid. The graphical plot updates each minute and tracks the AC Power and Daily Total throughout the day along with historical data from the best production day to date. The bar chart updates every 5 minutes and shows today's current production status against the totals for the prior two weeks.
The plot shows the instantaneous AC power which on an ideal clear day would appear as an inverted parabola coming away from 0 at sunrise, tracking through a maximum at the Sun's passage overhead and declining until sunset. It is not surprising that our best performance mimics this very closely. This optimum curve varies depending on the season as would our production potential. The scale for this is shown at the left of the plot. The production capacity of this array is 18.3 KW and the instantaneous power plot will not exceed that level.The veritcal bars indicate the typical desireable performance range for the time of year. This is taken from actual annual data for about a 3 week period centered around the current day. The low end of the bar is the seasonal average. For a better than average day we would want to be above that point and so there starts the range. The upper limit of the range is the maximum previously achieved for the seasonal period. The closer we are to that the better but if we were to exceed it, well, then we would be setting a record of sorts. So at a glance if the day falls within these ranges we're happy.
Cloud cover naturally impedes production. On a heavily overcast day it is not uncommon for the instantaneous power level to go as low as a couple of hundred watts or less. This varies greatly depending on the density of the cloud cover which itself will vary significantly during the course of the day. Production on a bright overcast day can still be significant and cover the energy requirements of the residence. Typically, however, our experience has been that on a wintry day we may generate only about 30-50 percent of our needs if that. Thankfully the Sun does peek out periodically often even on the worst of days.
Partly sunny or partly cloudy days can be particularly interesting. As passing clouds block the sunlight the instantaneous power levels drop abruptly and remain low depending on the duration of the shading. This depends on the size of the cloud of course and on how quickly it is moving. It is during these partly cloudy days that a very interesting effect occurs. Not only do the water droplets that form the clouds block the sunlight (and our view of the blue sky) they also serve as tiny lenses redirecting and reflecting it. As a result sunlight passing between the clouds can be a combination of direct light and refocused/reflected light. The resulting intensity on such days can be brighter than normal direct sunlight. This is evidenced by an increase in solar production with instantaneous power levels above the production curve (inverted parabola) that we see from a clear day. This is often pushing levels up to the 18.3 KW maximum capacity of the inverters. It is entirely possible that a partly cloudy day with just the right conditions might result in record production. We want those passing clouds to just miss the Sun, not block it at all, and redirect that sunlight our way.
About the Residence
The home uses energy both from electricity and from natural gas. Heating and cooling is supplied using a highly efficient two-stage geothermal heat pump. Water is circulated in five wells each drilled 185 feet down. We can optionally heat using a three-stage high-efficiency gas furnace. The gas furnace is used as backup for the geothermal system and also is utilized should the Grid Status fall into a deficit. It is the current opinion that it is more cost-effective to heat with the natural gas than electricity should we need to actually buy the electricity. Currently natural gas is also used to supply a fireplace insert, the kitchen stove, the water heater and the grill out on the deck. It is expect that the revenue from excess power generation will offset the gas purchased. We are effectively a Net Zero Energy Green Home.
Things of Interest
Since we are basically monitoring the sunlight reaching the property we can see the effect of things that might otherwise go unnoticed. We are situated just west of a set of ridges bordering the Laurel Highlands. The prevailing winds are typically from the west. The moist air along the ground is pushed upwards when it reaches the line of ridges. This tends to form clouds. It is this weather pattern that generates the added amounts of snow that the highlands enjoy during winter. As the Sun rises in the east the sunlight passes through this region of potential cloud cover. This often affects the power levels that we plot for an hour or two mid-morning. Look for dropouts and departures from the smooth curve.
When conditions are right aircraft aloft leave contrails. We tend to ignore these lines in the sky which, except for rare events, never appeared in skies until recently, probably only the past 50 or so years. We detect these on the plot as very brief slight dropouts. On an otherwise clear day it is very easy to make the correlation between the contrail and the disturbance on the plot. Generally this is not significant enough to affect the overall power generation for the day but it is interesting firstname.lastname@example.org