Tracer Data
 
Meteorological Data
 
Model Evaluation

 

Links

 

EPRI Bull Run Field Study

WHEN:

SF6 sampling is available for July 28, 1982 through October 18, 1982.

 

There were two "Intensives", when monitoring was enhanced and more detailed sampling was conducted. Intensive measurements were conducted over 53-week periods.The two "Intensives" were:

 

July 28, 1982 - August 29, 1982

September 20, 1982 - October 18, 1982

 

LOCATION: 

The Bull Run steam plant is located 20km west of Knoxville, TN and 8km east of Oak Ridge, TN, on the east bank of the Melton Hill Lake. Water from the Milton Hill Lake is used to cool the 950 MW, coal-fired, one-unit plant(twin furnaces operate as a single unit to product steam). The Bull Run steam plant is located in the broad Tennessee River valley (about 60 km wide) that separates the Great Smoky Mountains, rising to over 1700 m (msl) to the east and southeast, and the Cumberland Plateau, rising to over 900 m to the northwest. Within the broad Tennessee River valley, a system of smaller parallel ridges and valleys, with breaks in between, runs northeast-southwest. These minor ridges rise 60 to 180 m above the local valleys. In the immediate vicinity of the plant, the ridges are approximately 100 m high, with a distance of about 1 or 2 km between ridge tops. Most of the region is covered by forests. It is also important to note that the Melton Hill Lake, a 1 km-wide tributary of the Tennessee River, cuts perpendicular across the ridges near the Bull Run steam plant, which uses the river water for cooling.

UTM-E (km):            756.246

UTM-N (km):           3989.82

Latitude (deg):           36.0211 N

Longitude (deg):        84.1564 W

 

TRACERS:

Hourly Sulphur hexafluoride (SF6).

Hourly Sulfur Dioxide (SO2)

 

RELEASE:

All measurements of source emissions were made in the ductwork leading to the stack. The strip chart readout from the mass flowmeter was calibrated against timed weighings, several calibration points being taken each test day. Also available were rotometer readings. From the calibration data it was estimated that the flow rates taken from the mass flowmeter were accurate (at the 95% confidence level) to +/- 8.16 kg/hr, and the flow rates taken from the rotometer were accurate to +/- 4.08 kg/hr.

 

SAMPLING:

SF6 Tracer Monitoring Network

A network of approximately 1600 potential tracer sampling locations was used at the Bull Run site. The network design consisted of concentric circles at average radial distances of 0.5, 1, 2, 3, 5, 7, 10, 15, 20, 30, 40, and 50 km from the power plant. Using the existing roadway network, the downwind distance of the samplers assigned to an arc varied; the target accuracy was 1% of the distance from the power plant stack, or 50 meters, whichever was smaller. Tracer tests generally lasted six to nine hours. During this period, SF6 tracer gas was injected continuously into the ductwork of the stack. During each test approximately 200 sampling sites located on five to seven of the arcs in a sector ranging from 90o to 180o of arc, were operating. The sampling array was chosen on the basis of the expected meteorological conditions and remained fixed during any given test period

 

SF6 sampling is available for July 28, 1982 through October 18, 1982.

 

Hourly SO2 Monitoring Network

I have SO2 monitoring data only during the two "Intensive" data collection periods. There were a maximum of 20 receptors active for any 7-day period. The number of receptors and locations vary from 7-day period (168 hours) to the next. There were 10 7-day periods.

 

METEOROLOGY:

The meteorological measurements were collected from (1) a 122-m TVA tower (base elevation 317 m), located near the crest of a 70 m ridge about 1.8 km west of the stack, (2) the central observing station (base elevation 248 m), located at about the same elevation as the stack base in a field near the river about 5.3 km northwest of the stack, (3) one acoustic sounder (base elevation 259 m) about 1 km east of the stack, (4) various National Weather Service (NWS) stations. The attached map shows the locations of the TVA tower, central observing station (labeled as Micro-Met), and stack.

 

1. From the data collected onsite, I am distributing:

wind direction and wind speed from 10-, 30-m 50-, and 100-m

Temperature from 10-, 50-, and 100-m.

Dewpoint from 100m.

2-10m Delta-T

Atmospheric pressure,

Net, Solar and Total Sky Radiation

Acoustic Sounder Mixing Heights

 

2. From analyses conducted by the original EPRI investigators, I am distributing:

Monin-Obukhov length for dry and moist conditions

Surface friction velocity for dry and moist conditions

Mixing height estimates for dry and moist conditions

Pasquill Stability Category

Surface roughness length, albedo and bowen ratio as a function of wind sector and month

 

3. Rawindsonde observations from Nashville, TN.

 

4. Standard surface weather observations from six (6) NWS stations.

 

DISCUSSION:

The primary purpose of the EPRI Plume Model Validation and Development Project was to generate an extensive base of air quality and meteorological data. These data were to be used toward the first two objectives of the overall PMV&D Project:

 

1. Establish, by statistically rigorous procedures, the accuracy and uncertainty of ground-level concentrations predicted by existing plume models.

 

2. Assess model performance over a range of meteorological topographical and source conditions at a given site and determine the transferability of plume models from one site to another.

 

The field study site at the Kincaid Power Plant was chosen because it was isolated, so there would be no confusion as to where the tracer or SO2 was coming from, and because the terrain was relatively flat, and thus was free of terrain induced meteorological transport effects. The field study site at the Bull Run Steam Plant was chosen because it was isolated, so there would be no confusion as to where the tracer or SO2 was coming from, and because the rolling ridges in the vicinity of the release would provide a more challenging environment than that seen at Kincaid for characterization of the plume transport and dispersion.

 

As part of this effort, the data collected at Kincaid and at Bull Run was divided into two independent data sets so that the data used to evaluate an air quality transport and diffusion model would be independent of that used to develop a model. Half of the data set was used for a model development program sponsored by EPRI and was called the Developmental data set. The other half of the data set was used to evaluate the Hybrid Plume Dispersion Model (HPDM), Hanna and Paine (1989), and was called the Evaluation data set.

 

In 1991, a European initiative was launched for increased cooperation and standardization of atmospheric dispersion models for regulatory purposes. Conferences are held approximately every 18 months. As part of this initiative a Model Validation Kit [http://www.harmo.org/Kit/default.asp] was developed starting in 1993. Currently, the kit contains four field data sets as well as software for model evaluation. The Model Validatoin Kit does not have any of the Bull Run data at this time.

 

In this release of the data, the Bull Run Developmental and Evaluation Data Sets have been merged together, which essentially doubles the amount of data available for analysis.

 

The hourly SO2 data are being included so that research investigation can assess jointly the SF6 and SO2 sampling results.

 

DIGITAL DATA FILES:

Informational Files

 

BullRunDiscussion.pdf: Provides an overview of the data being made available. [42KB]

 

Logic_Behind_Kincaid_SF6-Arcs.pdf: This file describes why I was unsuccessful in combining the Developmental and Evaluation data sets and retain the subjective receptor placements used in the original investigations. I developed and used an objective scheme to place receptors along arcs that is nearly the same as that used in the original investigations. I also developed and used an objective scheme for assessing the quality of the SF6 for determination of the maximum concentration along an arc. [56.6MB]

 

KB48andKB49_SF6Data_Format.txt: Original archive format for SF6 concentration data values. [3KB]

 

BullRun-SF6-SamplingSummary.txt: In the original investigations, the data was divided into two (2) data sets; one for model development and one for model evaluation. This file list in a table format the days included in Developmental and Evaluation data sets. [3KB]

 

Bull Run Suspect SF604162012.pdf: In developing the Kincaid data for distribution, I developed an objective scheme for assessing the quality of the sampling along the SF6 arcs. I used the same objective scheme for the Bull Run SF6 data. This file discusses the 43 cases where my assessment scheme determined the maximum SF6 concentration might be suspect. Upon reviewing, it is my opinion that 5 cases are suspect; 28 cases are OK; 5 cases will be difficult to replicate (likely non-steady transport), and 5 cases fail to provide a representative arc-max due to truncated sampling along the arc. [152KB]

 

Note: as of 06192013 I have dedected another suspect max value, and will update this file accordingly, but it will take some time. The arc is 82 8 16 17 at 5km with a max of 3644 PPT, which is obviously erroneous. This arc looks reasonable, once you delete this erroneous value.

 

Meteorological Data Files

 

BullRunNearSurfaceMetData04162012.txt: Hourly near-surface meteorological observations collected at the Bull Run plant site. [226KB]

 

BullRunTowerMetData04122012.txt: Hourly meteorological observations collected from the 100-m tower located near the Kincaid plant site. [285KB]

 

BullRun-KB-88(NWS-UA5600).txt: US National Weather Service Upper Air data in 5600 format. Data begin August 26, 1982 and end October 25, 1982. Upper air balloon was released from Nashville, TN. [114KB]

 

NWS Hourly Surface Weather Data
Station Data File Description Size
Bristol.zip W13877.h82 W13877.txt [208KB]
Chattanooga.zip W13882.h82 W13882.txt [210KB]
Knoxsville.zip W13891.h82 W13891.txt [206KB]
Nashville.zip W13897.h82 W13897.txt [210KB]
Lexington.zip W93820.h82 W93820.txt [212KB]
Asheville.zip W038 12.h82 W03812.txt [207KB]

The hourly data for the stations listed above was obtained from: http://www.epa.gov/ceampubl/tools/metdata/us_met.html The meteorological data files available at this site contain measurements taken at 237 weather stations located throughout the United States for a period extending from 1961 to 1990. Exact collection dates vary by weather station. These data files are compatible with the EXAMS, PRZM, and EXPRESS systems distributed through the EPA Center for Exposure Assessment Modeling (CEAM).

 

SF6 Tracer Data Files

 

BullRunSF606192013.txt: Detailed listing of SF6 concentration values at each receptor along an arc. There were 12 receptor arcs, but not all were active during each hour of sampling. In this file, the data values are listed separately for each receptor arc that was active during each hour. Note, the receptors have been placed on arcs using an objective scheme that I developed when I was preparing the Kincaid data for distribution (see, Logic_Behind_Kincaid_SF6-Arcs.pdf). [6.6MB]

 

BullRunSF6Max06192013.dat: A listing of the maximum SF6 concentration seen along each arc and the objective ‘judgment’ code of whether the maximum is likely well characterized by the sampling available. I have also included the QC index for the Development Data distributed in the Harmonization Model Validation Kit, but since I used a different scheme for assigning receptors to arcs, these QC index values are only useful for informational purposes. [223KB]

 

BullRunSF6-Arcs06192013.pdf: This file displays the SF6 concentration values in a fashion similar to that employed by the original investigators. Note, the receptors have been placed on arcs using an objective scheme that I developed when I was preparing the Kincaid data for distribution. [534KB]

 

BullRun-KB-48(SF6-Developmental).txt: SF6 data for the Developmental data set, in the original format of the Bull Run data archive. [1.05MB]

 

BullRun-KB-49(SF6-Evaluation).txt: SF6 data for the Evaluation data set, in the original format of the Bull Run data archive. [923KB]

 

BullRunCombinedSF6-48and 49.txt: A spliced together version of the SF6 data for the Developmental and Evaluation data sets, in the original format of the Bull Run data archive. The format for this file and KB-48 and KB-49 is descrbed in KB48andKB49_SF6Data_Format.txt. [1.69MB]

 

Hourly SO2 Data Files

 

BullRunHourlySO204112012.txt: Listing of SO2 concentration values (PPB) at each receptor for each hour. There were a maximum of 20 receptors active for any 7-day period. The number of receptors and locations vary from 7-day period (168 hours) to the next. There were 10 7-day periods. [325KB]

 

BullRunSO2Format.txt: Description of format used to create data archive of hourly SO2 values (BullRunHourlySO204112012.txt). [4KB]

 

BullRunListSO2SourceDatao4172012.txt: I processed the data archive files available to me, in order to derive an SO2 emission rate, so that the hourly SO2 monitoring data might be used in model performance analyses. Unfortunately, I could only compute SO2 emission rates for the 480 period from Aug 1, 1982 through Aug 20, 1982. During this 480 hour period the maximum emission SO2 rate was 2786 g/s; the minimum SO2 emission rate was 1463 g/s, and the average SO2 emission rate was 2035 g/s.

 

REFERENCES:

Hanna, S.R., and Paine, R.J., (1989): Hybrid plume dispersion model (HPDM) development and evaluation. J. Of Applied Meteorology. (28):206-224.

 

Hudischewskyj, A.B., and Reynolds, S.D., (1985): Catalog of Data for the EPRI Plume Model Validation and Development Data Base: Moderately Complex Terrain Site. Prepared for the Electric Power Research Institute Environmental Physics and Chemistry Program, Energy Analysis and Environment Division, EA-3762, Research Project 1616-9.

Acknowledgements:

I have reformatted the data that was provided to me. If there are errors in the data I have provided, most likely they are a result of my efforts to format and organize the data for distribution. No fault should be placed on the kind people who provided me with their data files.

Joseph C. Chang provided me with all of his Bull Run data files, which we have come to understand were likely created by someone working for SAI, many years ago. Gary Moore provided me with all of his Bull Run data files, which are the ten archive one-week files I discuss.

Through a series of Emails, Norman Bowne, Roger Brode, Steve Hanna, Jayant Hardikar, Russ Lee, Douglas R. Murray, Helge Olesen, Bob Paine, James Paumier, Don Shearer, and Dave Strimaitis have all tried to help me decipher the avaliable data files. As you can see, I need lots of help, and I am honored that so many have offered their help.