[ANSS-netops] solar power problems

Wed Jan 7 15:29:10 UTC 2015

Hi Mitch et al.
This is a good discussion of Power for seismic stations, thanks.

In the Transportable Array deployment, all 1717 stations operate(d) off 
solar power for at east two years.  The average current draw for a 
station is 0.5Amps on a 12V system, but can vary from 0.4 to 0.6A 
depending on the telemetry system. High current telemetry systems such 
as VSAT are powered separately.  We avoided AC power because of the 
proximity of noise due to pumps, motors, etc.

In general we use (2) 90W PV panels and (2) AGM Lead Acid batteries for 
stations south of the Kentucky/Tennessee line (row U in TA station 
codes) and (3) 90W PV panels and (3) AGM Lead Acid batteris north of 
there. Shady or snowy sites occasionally got more panels and batteries.  
For permanent stations I'd go with the (3) PV and 3 or 4 batteries.  We 
prefer good quality batteries designed for solar applications, such as 
the Concorde Sun Extender PVX-1040T.  We use PWM regulators with Low 
voltage disconnect at 10.8V.   A few more sophisticated options are 
discussed below.

I would concur with the notion that the most effective way to improve a 
marginal station power situation is to add 1 or 2 batteries, and often 
this can be done without much infrastructure alteration. And the next  
option is to add a panel.  There is little concern about over driving 
the charge controller with too much current from too many panels.  In 
Alaska, the "more batteries" approach is taken to extremes in which 
stations have 24 batteries to float through the winter. I would also 
concur that, to date, neither wind nor fuel cells have proved reliable 
enough to warrant their use, especially in permanent stations of the 
Lower48.

More complexity described below:
Our system has, in addition to the main battery bank, a small reserve 
battery.  When the system switches to the reserve, certain loads such as 
the telemetry radio and local data storage are duty cycled at four hour 
intervals.  This reduces the power of the station to about 3W, yet still 
provides complete telemetry (though with episodic latency) and complete 
local storage.  For us, this reserve power serves to identify the source 
of the outage is clearly power as opposed to a host of other 
possibilities.  In the original design this reserve battery was Alkaline 
Lattern 30AH batteries [(3) x 6Volts] (a primary battery, disposed of 
after use).  More recently, We have also used 100-300AH rechargeable 
batteries that are then connected to the main batteries using a battery 
isolator circuit-which connects the reserve batteries to the charger 
only when the main battery has recharged to 13.2V.  We add a 10A current 
limit to the battery interconnection.  The reserve power load shedding  
can be thought of as doubling the capacity of the reserve batteries, 
reducing the cost of overall power system for this reserve capability.  
Without that sophisitication of load shedding, adding more batteries is 
effective but there is a cost in terms of station uptime. When a very 
large, undifferentiated battery bank is depleted, it will take a longer 
time for the batteries to reach the reconnect voltage.  In this time, 
the station itself could be operating on the minimal power produced. We 
keep the main battery bank fairly modest so it recovers voltage quickly, 
and defer recharging the reserve pack until there is ample 
power-sometimes weeks, or in Alaska, months later.  Its meant to get you 
through an ocasional bad spell.  One issue in this reserve battery 
switching is the dc currents can introduce magnetic pulses seen on the 
seismometer-particularly Trilliums within a few meters of the switches.

If you are plannning a Net-ops meeting in the future, I'd be happly to 
elaborate on the power system for Alaska, which uses a Genasun MPPT 
charge controller and LiFePO4 (Lithium Ion) batteries and the same duty 
cycle loads, reserve battery concepts.  These significantly reduce the 
weight of a 1440AH system to 420 lbs and do not require derating the 
capacity for cold temperatures.  They are very expensive, but not as 
much as a helicopter trip.

Bob Busby
TA Manager

On 1/6/2015 7:36 AM, Kyle Persefield wrote:
> Mitch,
>
> My 2 cents worth
>
> Because of cost, we have found throwing on more solar panels to be the
> cheapest and least maintenance intensive solution.  Fuel cells and
> thermoelectric generators are expensive and then there is the recurring
> cost for fuel, getting fuel to the site, then monitoring of the fuel
> supply level to consider, and the added required maintenance.  We have not
> found a solution to use these devices as demand requires their use.  Or
> turning them on and off as needed.  So long as there is fuel they are on.
>
> We have been very disappointed with wind turbines.  The smaller ones,
> which are designed for the consumer market, the bearings always fail.
> Expect no more than 2 or3 years out of these "cheap" units.  Then of
> course there is the need for wind.  No wind for extended periods is just
> as bad as your overcast scenario.
>
> Kyle
>
> -----Original Message-----
> From: ANSS-netops [mailto:anss-netops-bounces at geohazards.usgs.gov] On
> Behalf Of Mitchell M Withers (mwithers)
> Sent: Monday, January 05, 2015 7:12 AM
> To: Philip Crotwell
> Cc: anss-netops at geohazards.usgs.gov
> Subject: Re: [ANSS-netops] solar power problems
>
>
> Here is a snippet from a recent report from a visit to an example station
> with a reftek, three S-13's, and an episensor.  Stations vary of course
> and we do use low voltage cutouts at every station (fancy ones that cut
> out the transmitter first, then the DAS and everything else if the voltage
> continues to get lower).
>
> "The new battery banks, when installed were at 12.95 and 12.98. The total
> station draw is exactly 600ma (checked continuously for about three
> minutes). I did a quick calculation of 4 batteries at 96AH each, 384/.6 =
> 640/24 =26.6 days. This calculation would assume no solar charge, but does
> not take into account reduced battery capacity due to cold temperatures."
>
> The panels at that particular station were supplying about 700ma together
> on an overcast day and are being replaced with bigger panels this week.
> Of course one solution is more battery and more solar at every station
> along with more frequent refreshing of batteries.  But that gets expensive
> and time consuming so I was fishing to see if anyone is doing something
> creative. (e.g. wind or hamster wheels).
>
> Mitch
>
> Center for Earthquake Research and Information (CERI)
> University of Memphis                Ph: 901-678-4940
> Memphis, TN 38152                   Fax: 901-678-4734
>
>
> ________________________________________
> From: Philip Crotwell <crotwell at seis.sc.edu>
> Sent: Monday, January 5, 2015 7:50 AM
> To: Mitchell M Withers (mwithers)
> Cc: anss-netops at geohazards.usgs.gov
> Subject: Re: [ANSS-netops] solar power problems
>
> Hi
>
> How did you come up with your 25 day figure? Can you put some numbers on
> power input and output?
>
> We use two 105 amp-hour batteries per station, where the load is about
> 1/2 an amp. That gives me about 17.5 days theoretically, but my
> understanding is that you never want to discharge batteries anywhere near
> their rating as they can be damaged by high discharges. So maybe worry
> less about age and more about installed capacity, ie double the battery
> and replace them half as often.
>
> We also, because of the cell modems, can monitor the battery voltage over
> time, we have a cron job to ping the cell modem once an hour and ask it
> what the input voltage is. For example here is the last few days at one
> station. You can definitely tell the difference between sunny days and
> rain, and we get a heads up if the power is getting low and can do
> something before the station goes down.
> http://eeyore.seis.sc.edu/earthworm/status/HAW_last720.png
>
> Here is another station that we are becoming worried about, looks like I
> might get to go on a road trip soon!
> http://eeyore.seis.sc.edu/earthworm/status/CASEE_last720.png
>
> Philip
>
> On Sun, Jan 4, 2015 at 9:07 AM, Mitchell M Withers (mwithers)
> <mwithers at memphis.edu> wrote:
>> Many of our stations run on battery and solar and that normally works
> well.  We have a routine battery replacement cycle to make sure they don't
> get old.  Theoretically, we should be able to run with zero solar for
> about 25 days.  But this has been an unusually dreary winter in the
> southeast and we haven't had much sun for the past two months or more.
> I'm wondering what others do in areas with limited sunlight to power
> stations that don't have AC available?
>> Mitch
>>
>> Center for Earthquake Research and Information (CERI)
>> University of Memphis                Ph: 901-678-4940
>> Memphis, TN 38152                   Fax: 901-678-4734
>>
>> _______________________________________________
>> ANSS-netops mailing list
>> ANSS-netops at geohazards.usgs.gov
>> https://geohazards.usgs.gov/mailman/listinfo/anss-netops
> _______________________________________________
> ANSS-netops mailing list
> ANSS-netops at geohazards.usgs.gov
> https://geohazards.usgs.gov/mailman/listinfo/anss-netops
> _______________________________________________
> ANSS-netops mailing list
> ANSS-netops at geohazards.usgs.gov
> https://geohazards.usgs.gov/mailman/listinfo/anss-netops
>

-- 

============================================================

Robert W. Busby
Transportable Array Manager           508-801-7628
USArray / EarthScope                  37 Haynes Avenue
www.earthscope.org/usarray            Falmouth MA USA 02540-2312