May 15, 2014

POINT OF CONTACT

Principle investigator:
     Kirk L. Clawson
     NOAA Air Resources Laboratory Field Research Division
     1750 Foote Dr.
     Idaho Falls, ID 83402
     Kirk.Clawson@noaa.gov
     (208) 526-2742

README.TXT - Description of SF6 Release and Data Files

THE SF6 TRACER RELEASE SYSTEM

The SF6 tracer release system was custom built for PSB1 at the ARLFRD 
office in Idaho Falls, ID. The system was placed in a cargo trailer to simplify 
deployment, provide a reasonably controlled environment for operation, and to 
simplify removal of the release system when the field deployment was complete. 
The complete release system was entirely self-contained in the cargo trailer and 
only required a 155 VAC 20 ampere power source. This was provided from an 
adjacent power pole.

The ARLFRD tracer release system was engineered to release a constant amount 
of SF6 from a single point source at the center of the Grid 3 tracer facility. 
Each SF6 point source release during PSB1 lasted a total of 2.5 hours. The 
first half-hour of each release period was dedicated to obtaining steady-state 
dispersion conditions over the entire sampling area before sampling began. Each 
release then continued at the initial release rate for the next two hours for 
the actual 2-hour long tracer sampling period.

During all SF6 releases, the gaseous tracer flowed from a cylinder containing 
SF6 through the mass flow controller, through a visible flow meter, and into a 
garden hose. The outlet end of the garden hose served as the dissemination point. 
The garden hose outlet was placed at a height of 1.5 m AGL attached to a tower 
at the center of the sampling grid. It was oriented horizontally to avoid 
imparting any vertical momentum to the tracer. A heater was used to maintain 
constant pressure in the SF6 cylinder and to assist with the vaporization of 
the liquid SF6. The SF6 tracer was provided in K-size cylinders by Norco, Inc. 
The certified concentration of the liquid SF6, as reported by Norco, was >99.9%.

The heart of the SF6 tracer release system was the thermal mass flow controller 
(Hastings Teledyne, Model HFC-203). The mass flow controller was responsible 
for monitoring and controlling the tracer leaving the SF6 cylinder. During a 
release, a voltage was applied to the mass flow controller that was proportional 
to a given SF6 flow rate. This voltage could be manually controlled to obtain 
any desired release rate between a set range. The voltage and the flow rate 
from the mass flow controller were continuously monitored and recorded with a 
datalogger.

Accuracy

The mass flow controller was calibrated at the factory and subsequently 
double-checked outdoors at our office in Idaho Falls. Calibration was needed to 
correlate the tracer flow rate to the applied voltage. Several verification tests 
were conducted after the factory calibration to ensure proper functioning of the 
mass flow controller.

SF6 Release Summary

A total of 5 tracer tests (IOPs) were conducted. The releases were always 
conducted from the center of the sampling grid. The target SF6 release rates 
for the first three IOPs was 10.0 g/s. This release rate was selected in order 
to provide tracer concentrations sufficiently high enough to be sampled by the
real-time analyzer installed in the aircraft. The aircraft was used for sampling 
the vertical distribution of the tracer plume during the first three IOPs but 
was not available for the last two. A much lower release rate was used for the 
remaining tests. The target SF6 release rate for these tests was 1.00 g/s.

Target and actual SF6 tracer release rates are listed in the table below. 
In the table, MFC is the flow rate measured by the mass flow controller and 
Scale is the difference in mass of the SF6 cylinder between the start and end 
of the release. Actual release rates differed only slightly from the target 
release rates. For the first three tests, the actual release rates ranged from 
9.93 to 10.18 g/s, which were anywhere from 0.7% under to 1.8% greater than 
intended. For the last two tests, the actual release rates ranged from 1.030 to 
1.043 g/s, which were anywhere from 3.0% to 4.3% greater than intended. The 
standard deviations of the actual flow rates for the first three tests ranged 
from 0.01 to 0.26 g/s. Standard deviations of the actual flow rates for the last 
two tests ranged from 0.007 to 0.010 g/s. The low standard deviations indicated 
very steady flow rates throughout the entire 2.5-hr continuous release periods. 
The total amount of SF6 tracer material that was disseminated during field deployment 
was 289,905 g.

					Total	Total
			Release		SF6	SF6
			Start	End	Release (g)	Correction	Target	Measured	S.D.
IOP	Day	Month	HHMM (MST)	Scale	MFC	(Scale/MFC)
1	02	Oct	1400	1630	91600	89049.1	1.0286		10.0	10.177		0.26
2	05	Oct	1230	1500	89900	89509.2	1.0044		10.0	9.986		0.20
3	07	Oct	1230	1500	89400	89604.6	0.9977		10.0	9.930		0.01
4	11	Oct	1330	1600	9525.4	9063.93	1.0509		1.0	1.043		0.01
5	18	Oct	1230	1500	9480	9030.92	1.0497		1.0	1.030		0.007	

Graphs showing release rates and the cumulative amount of SF6 tracer released 
during each IOP indicate very steady release rates, with the exception of IOPs 
1 and 2. At about 15:18 and 13:50 MST for IOPs 1 and 2, respectively, the source 
of the SF6 tracer was switched from a near-empty bottle to a full bottle. While 
the trace indicates a large flow excursion, this is an artifact of the mass flow 
controller. Visual monitoring of the rotometer showed only a minor and very short-lived 
flow disruption. For all practical purposes, the flow rate remained 
steady. This is also indicated by the cumulative trace of disseminated SF6 tracer, 
which shows only a steady increase for those two tests.

SF6 Release Quality Control

The quality control program for the SF6 tracer release consisted of the 8 steps 
outlined below:

1. Pre-project preparation.
2. Pre-test procedures.
3. Monitoring of key operational parameters during the test.
4. Post-test procedures.
5. Post-test data screening and processing.
6. Verification of all calculations and data by a second analyst.
7. Identification of data problems and setting of QC flags.
8. Review of final data files.

1. Pre-project preparation.

Before the experiment, the SF6 release mechanism was constructed and thoroughly 
tested to ensure all systems were in good working order. Prior to the release 
system construction, the mass flow controller was calibrated at the factory and 
again at the FRD office/laboratory facility to correlate the actual flow rate 
with the indicated flow rate. After construction, the system was tested from end 
to end for flow accuracy.  The release system released 99.9% pure SF6 without dilution. 

2. Pre-test procedures.

On the day of a test, the release system operator was required to follow 
established procedures for preparing the release mechanism. These procedures 
were based on the experience of previous tracer projects. The procedure included 
checking for loose connections, visually inspecting the release line, 
calibrating the scale, setting the clock, setting the mass flow controller output 
to zero, and verifying that data was being recording on the computer. These 
actions were recorded in the release logbook.

3. Monitoring of key operational parameters during the test.

During the test, the mass flow controller and weight of the SF6 bottle were 
monitored for a stable and correct flow rate. These values were recorded 
approximately every 10 minutes in the release logbook. The release system 
operator was able to adjust the flow rate on the release mechanism if necessary. 
Note: The mass flow meter was accurate enough that it did not require additional 
adjustment after initial setting at the beginning of each test.

4. Post-test procedures.

After a test was complete, the release system operator followed end of release 
procedures for shutting down the release mechanism and collecting the data. Weight 
loss from the SF6 bottle(s) was recorded in the release logbook. Release data that 
had been recorded on the computer was backed up onto a compact memory stick and 
returned to the FRD office for processing.

5. Post-test data screening and processing.

Once the memory stick was returned to FRD, the data was uploaded onto the 
network for processing. Release rate data was graphed and reviewed for any 
spikes or anomalies in the recorded data that would indicate deviations from 
a stable flow rate. Release rate data from the mass flow controller was 
compared to the actual weight of the released tracer, as measured by the scales, 
to ensure that the flow rate was within five percent of the mass flow set point. 
The mass flow output data was adjusted (corrected) to match the total amount 
released using the precision balance scale data.

6. Verification of all calculations and data by a second analyst.

The plots of the new data were reviewed and verified by a second analyst.

7. Identification of data problems and setting of QC flags.

The release logbook entries and the plots of the data were carefully reviewed 
by the data analysts. No problems were found. If any problems had been found, 
they would have been annotated with the correct flag and recorded in the final 
data files. The data flags would indicate unstable or varying flows, spikes in 
the release rate, or missing data.

8. Review of final data files.

The data files were carefully reviewed for any problems and checked for the 
correct flags.

Data File Format

The one second readings from the mass flow controllers are provided in data 
files on the CD accompanying this report. The files are named ReleaseX_PSB1.csv, 
where X is replaced by the IOP test number. The files contain five columns:

1. date (month/day/year)
2. time (hhmm in MST)
3. seconds
3. tracer flow rate (grams per second)
4. quality flag

The files are all comma separated variable format. The first line of each 
file contains headers for each column. Quality flags are 0 for good data, 1 
for suspect data.
