Belmont Water Department
35 Woodland Street
P.O. Box 56
Belmont MA 02478
.PP
This is to acknowledge receipt of your bill of $651.59
for water-sewer charges for the period of
August 11, 2004 to November 17, 2004 for the
above captioned account.
For reasons set out below, I believe these charges to
be in error. Relying on the instructions printed on
the reverse side of each water-sewer bill of the 
proper procedure in the event of meter failure,
I have calculated what I believe to be an equitable
charge, and I respectfully request that you correct
my account accordingly.
.PP
I respectfully draw your attention to the table below,
all but the last line abstracted from your bills
which lists the water consumption for the respective
billing periods and calculates the rate of at which
water was used in cubic feet per day. The meter reading
for today, December 10, 2004 was made by myself.
.sp
.nf
08-16-2003 to 11-19-2003 : 1300 cu ft in  84 days = 15.476  cu ft/day
11-19-2003 to 02-14-2004 : 1810 cu ft in  86 days = 21.047  cu ft/day
02-14-2004 to 05-10-2004 : 1658 cu ft in  85 days = 19.506  cu ft/day 
05-10-2004 to 09-15-2004 : 1910 cu ft in 127 days = 15.039  cu ft/day 
09-15-2004 to 11-17-2004 : 6138 cu ft in  62 days = 99.000  cu ft/day
11-17-2004 to 12-10-2004 :  531 cu ft in  22 days = 24.136  cu ft/day
.PP
The foregoing table shows five readings over a span of 404 days
in the range of 15.039 to 24.136 cu ft/day, giving an average
consumption for these periods of 19.041 cu ft/day, and one period
of 62 days from 9-15-2004 to 11-17-2004 in which water consumption
appears to have skyrocketed more than five fold to 99 cu ft/day.
This figure is all the more remarkable in consideration of the fact
that within that 62 day period, there were 23 days and 21 nights
when 174 School Street was vacant, leaving only 39 days in which
6138 cu ft of water could have been knowingly consumed.
(From September 27 through November 30, 2004, and from November 15
through November 17, my wife and I were on Nantucket, and on
October 18 we left for a trip to Virginia, returning on November
3, 2004.)  Moreover during the 62 day period, no running faucets,
no malfunctioning toilets were noted, and no plumbing repairs
of any kind were made either then or subsequently.
.PP
Turning now to the actual bill, of which I attach a copy, please
note that two separate meters, No 21721329 and No 90657390 are invoked.
One of your staff offered me the explanation, if I understood correctly,
that meter No 21721329 had been relied upon for measuring
the consumption from August 11 2004 to September 15, 2004,
on which date it was found that the actual reading of this
meter (5378 cu ft) was 570 cu ft less than the estimated
"reading" (5948) on August 11, 2004, that I had in effect
prepaid not only the entire consumption through September 15,
but that I had also paid for 570 cu ft not yet consumed
a charge which was credited against subsequent use. It was further
explained to me, if I understood correctly, that the second meter,
no 90657390 was newly installed on September 15, 2004,
that since it was a new meter its reading was known to be 0,
and with such assurance that this reading did not require
to be recorded on the bill. The new meter no 90657390 was
first read on November 17, 2004, electronically so that
a mistake was impossible, and showed that in the preceding 62
days we had consumed 6138 cu ft of water, incurring a debt
of $650.65, net of the 570 cu ft with which we had been credited.
Your staff also assured me that there could be no question
that 6138 cu ft of water "had gone through the meter." 
.PP
I don't believe it. In the first place, I have no memory
of the water meter having been replaced on September 15, 2004.
But then I may be wrong. I am seventy four and a half years
old, and my memory is not what it once was. I believe this
water meter was changed about a year earlier (November 19,
2003). At that time a "Badger" module (No 67497204) which
had been mounted on the outside of the house was disconnected.
The bill of that date lists a reading of 93700 cu ft, and
this evening the meter still exhibits that value 93704 cu ft,
to be exact. Furthermore the meter no 90657390 refers,
so far as I can tell, to the large globular brass fitting
with the number 90657390 stamped on its body. This number
is already registered on the earliest water bill I can
conveniently locate, dated 05-03-2001. Hence, the representation
that meter no 90657390 was newly installed on September 15,
2004 with an indisputable reading of 0 cu ft, is unpersuasive.
Device no 90657390 as it is ptresently configured,
appears to contain not a recording device 
but merely the vanes that drive a separate recording meter.
In its present configuration, 90657390 
is not capable of displaying or transmitting any reading at all.
Device no 90657390 is coupled mechanically to a  combined
meter-telemtry module labeled with a bar code and
the number 21721329.
.PP
Hence the representation that meter 21721329 was replaced on
September 15, 2004 is similarly unpersuasive.
21721329 is still in place;
a bar-coded label betrays its identity.
I infer that it was installed on November 19, 2003,
was properly set to 0 cu ft at that time
and has been relied on to transmit telemetry readings.
For whatever reason, these telemetry readings appear
unreliable. The telemetry reading (ERT) of August 11, 2004
was about 1270 cu ft (63.5 days'consumption) in error.
(Assuming 20 cu ft/day consumption and an actual reading
of 5378 on 9/15/04, 35 days previously, on 08/11/04,
the true reading would have been 5378 - 700 or 4678
instead of 5948 which it in fact registered.)
.PP
The mistake in the bill now becomes obvious.
The three meter readings
.nf
 8/11/04 5948 ERT
 9/15/04 5378 ACT
11/17/04 6138 ERT
.PP
are all output of the same instrument,
the 21721329 telemetry meter on a 90657390 base.
The data transmitted by the 21721329 telemetry meter 
have proved inaccurate in the past,
and will most likely prove inaccurate again.
But the 9/15/04 reading was actual and reliable;
and even though the 11/17/04 reading was telemetric
it is, for the time being appropriately relied upon.
The the billing error is not from the meter.
It is a very simple and a very serious accounting error.
.PP
The difference of 570 cu ft by which the 
8/11/04 reading exceeded the 9/15/04 reading
was properly credited. But, furthermore,
if on 9/15/04 an actual reading of 5378 was found,
then this actual reading may not
be ignored on the false pretext that the meter
had been changed: it had not.
To ignore the actual 5378 cu ft reading
and arbitrarily substitute 0 cu ft,
with the explanation that a faucet must have been dripping
or a toilet plunger not seated properly
and to assert categorically that 
6138 cu ft of water have flowed through the meter
is an unacceptable alibi for a serious clerical error.
.PP
The 9/15/04 5378 ACT and the 11/17/04 6138 ERT readings
must be the basis on which the consumption between 9/15/04
and 11/17/04 is calculated.
Then, if,
for the purposes of the present billing,
one trusts the 11/17/04 6138 ERT reading,
the consumption in the period
between 9/15/04 and 11/17/04 will be 6138 - 5378
or 760 cu ft. Accepting the claim that the house
was occupied only 39 days of this period one calculates
a consumption of 19.847 cu ft/day, consistent with
the average consumption of 19.041 cu ft/day calculated
from prior usage.