Page 402 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

Details of assumptions and calculations used in QRA 1377
Lake travis drinking water contamination flat, high permeability range lands in the western portion of
(Pedernales Watershed) the pipeline.
A number of river and stream crossings in the Pedernales water- Prime agricultural land contamination
shed were rated as hypersensitive for potential drinking water
quality impacts to Lake Travis. A spill volume of 500 bbl is set as the threshold for impacts
Additional creeks as well as some dry channels identified as to agricultural lands. A spill this size resulting from a rup-
potential overland flow paths of concern, were identified as ture could be expected to contaminate about 114 of an acre of
sensitive. The total mileage of these sensitive (Tier 2) and soil.
hypersensitive (Tier 3) stretches along the pipeline were fac- Impacts to agriculture were evaluated by reviewing soils data
tored in as locations which could impact Lake Travis water from U.S. Department ofAgriculture databases. Prime agricul-
quality, using the factors described for “Drinking Water tural was identified as those farmlands having the following
Contamination’’ in this section. soil types: BaA, BaB, BeB, Bo, BuB, HeB. HOB, KrA, Nd, No.
ROB, Sa, Sg, Sm, andTr.
Recreational water contamination The distance of these types of soils crossed by the pipeline
was measured with the supposition that any prime farmland
The potential for recreational waternays contamination is along the pipeline could be impacted from apipeline accident
based on the idea that any product spill which reaches a water- up to a distance of 1,250 ft from the point of release.
way has the potential for negatively impacting recreational Therefore, the band of impact along the pipeline for evaluat-
uses. This may be a result of short-term impacts to surface ing any point was 2,500 ft. In most cases, overland spread
water quality which limit contact recreation, and fish kills or would cause impacts of two to three acres from any individual
contamination which may limit recreational fishing. spill event.
Two thresholds of spill size were used in determining Although localized channels, ditches, or roadways may pro-
whether a surface water body would potentially be affected by a vide a conduit for product to avoid major contamination of
spill. For portions of the pipeline where it is more likely that a farmland in general, it is assumed that any release over farm-
spill would impact a surface water body, a threshold of 500 bbls land will have an impact to that farmland. Therefore, a proba-
was used. For those portions of the pipeline that were either bility of 100 percent for impacts to agriculture is associated
very remote from the potentially threatened surface water body, with any release over prime farmland.
or which were in an area of very flat topography, a threshold of For most ofthe pipeline, it was assumed that prime farmland
1500 bbls was used as a minimum spill size. was over Tier 1 areas. However, in Bastrop County, where a
It should be noted that most ofthe streams that are crossed by major portion of the pipeline is rated as sensitive for potential
the pipeline are small, and in many cases are seasonal. A prod- contamination of ground water resources, the distance of agri-
uct release may therefore result in a large portion of the total cultural lands covered by Tier 1 and Tier 2 portions of the
stream flow consisting of product contaminants, for some dis- pipeline were tabulated separately.
tance downstream from the point ofrelease. Therefore, aproba- The average farmlands crossing distance was 872 ft, and the
bility of 100 percent for contamination was set for any median 94 ft.
100-meter segment along the pipeline containing a river or
stream crossing as well as for each of the adjoining 100-meter Wetlands contamination
segments in order to account for the close overland pathways
which could impact a stream. A spill volume of 500 bbl is set as the threshold for impacts to
In addition, some probability exists that a release at addi- wetlands.
tional points in the watershed may impact the surface water Two separate types ofwetlands crossings are noted along the
quality. Since overland flow modeling was performed to iden- pipeline right-of-way-palustrine and riverine. A total of 967
tify the flow pathways from points along the pipeline, the char- wetland areas were identified within the pipeline corridor, with
acteristics of these flow pathways were used to establish for a total of 159.7 miles of pipeline crossing or adjacent to wet-
each pathway a probability of impacting the surface water lands. These figures were tabulated by comparing the pipeline
stream during a major release. right-of-way with national wetlands inventory maps. Of the
These characteristics included distance from the pipeline wetlands types, there were 857 palustrine wetlands which could
along the pathway to the surface water body, slope ofthe path- be potentially impacted, consisting mainly of small ponds
way, terrain type (urban, agricultural, forested range land)- within the 2,500-foot (ft) corridor. The average linear distance
as an indicator of ground cover which could promote or retard of the palustrine wetlands is 711 ft. The average linear
overland flow, and soil permeability. These characteristics are distance of the 110 riverine wetlands is 2,127 ft. with a
used to generate a composite number for each flow pathway. median distance of 1,339 ft.
Those pathways which were not within a 300-meter band Therefore, the potential for impact to any wetland resource is
across each stream crossing, but which had a score equal to or represented by the distance across the wetland plus 1.250 ft to
higher than the 300-meter band were assigned a probability either side along the pipeline. A length of analysis for impacts
of impact of 90 percent. Areas of lower scores were rated to individual wetlands is set at 3,372 ft in order to encompass
incrementally with probabilities of 70 percent and 40 percent. the average wetland crossing, plus the 1,250 ft to either side
The final two sets ofpathways were scored at 10 percent and 0 which could impact the wetland during a spill. The probability
percent probability. Pathways that are assigned a 0 percent of impact from a spill into or proximal to the wetland is set at
probability largely represent points along the pipeline over 100 percent.