Final TMDL
TOTAL MAXIMUM DAILY LOAD (TMDL) ASSESSMENT Terrace Reservoir Conejos County, Colorado January 2013 TMDL SUMMARY
Waterbody Name/Segment Number
Main Terrace Reservoir; CORGAL08
Pollutant/Condition Addressed
Total Recoverable Iron
Affected Portion of Segment
All
Use Classification
Aquatic Life Cold 2 Recreation E Agriculture
Waterbody Designation
Use-Protected
Water Quality Target
1000ug/L (Total recoverable)
TMDL Goal
Attainment of the assigned use standards for total recoverable iron.
EXECUTIVE SUMMARY Alamosa River Segment 08, Terrace Reservoir, was added to Colorado’s 303(d) list of water quality impaired waterbodies for nonattainment of water quality standard for total recoverable iron in 2008 (Table 1). Excess total recoverable iron (Fe-Trec) impairs the Aquatic Life Cold 2 classification for Segment 08. The high concentration of this metal is primarily the result of the natural geologic conditions of the area and historical mining activity in the area. Terrace reservoir is an on-channel reservoir on the Alamosa River located in Conejos County in south-central Colorado. The goal of this TMDL is to identify the reductions in Fe-Trec loadings necessary to attain the assigned numeric water quality standard and thereby support the designated uses. [1]
Final TMDL
Segment # Segment 08
Segment Description Terrace Reservoir
Portion all
303(d) Listed Contaminants Total recoverable iron
Table 1. Segment within the Alamosa River watershed that appears on the 2008, 2010, and 2012 303(d) lists of impaired water bodies.
.
Figure 1. Terrace Reservoir. II. INTRODUCTION Section 303(d) of the federal Clean Water Act requires states to periodically submit to the U.S. Environmental Protection Agency (EPA) a list of water bodies that are water-quality impaired. A water-quality impaired segment is a segment that does not meet the standards for its [2]
Final TMDL
assigned use classification. The list of impaired water bodies is referred to as the “303(d) list”. The list is adopted by the Water Quality Control Commission (WQCC) as Regulation No. 93 (WQCC, 2012). For water bodies and streams on the 303(d) list, a Total Maximum Daily Load (TMDL) is used to determine the maximum amount of a pollutant that a water body may receive and still maintain water quality standards. The TMDL is the sum of the Waste Load Allocation (WLA), which is the load from permitted and non-permitted point source discharge, Load Allocation (LA) which is the load attributed to natural background and/or non-point sources, and a Margin of Safety (MOS) (Equation 1). (Equation 1): TMDL=WLA+LA+MOS Alternatively, a segment or pollutant may be removed from the list if the applicable standard is attained, if implementation of clean-up activities via alternate means will result in attainment of standards, if the original listing decision is shown to be in error, or if the standards have been changed as the result of a Use Attainability Analysis (UAA), or other EPA approved recalculation method. Terrace Reservoir is included on the 2008, 2010 and 2012 303(d) lists for exceeding the Aquatic Life Use-based chronic standard for Fe-Trec (Table 1). This Aquatic Life Use is the only assigned use on Terrace Reservoir that includes an iron standard. The impairment status for designated uses in Terrace Reservoir is presented in Table 2. The other assigned uses are not impaired.
Date (Cycle Year) of Current Approved 303(d) list: 2012 WBID Segment Description Designated Uses & Impairment Status CORGAL08
Aquatic Life Cold 2: Impaired Recreation E: Not Impaired Agriculture: Not Impaired
Terrace Reservoir
Table 2. Designated uses and impairment status for Terrace Reservoir, segment CORGAL08.
III. GEOGRAPHICAL EXTENT Terrace Reservoir is part of the Rio Grande River Alamosa River Basin, Hydrologic Unit Code (HUC) 13010002. Terrace Reservoir is an irrigation reservoir located on the Alamosa River and is located in Conejos County. Terrace Reservoir is located in south-central Colorado in the San Juan Mountain Range about 25 miles south-west of the Town of Alamosa. Terrace Reservoir is owned and operated by the Terrace Irrigation Company. The reservoir was constructed in 1912, and had a surface area of approximately 300 acres. The National [3]
Final TMDL
Hydrography Dataset (NHD) reports a surface area of 142 acres. The reservoir volume fluctuates with the irrigation season, and ranges from a maximum of 13,000 acre-feet to a minimum of about 2000 acre-feet. The reservoir is approximately 14 miles downstream of Wightman Fork. Further upstream (approximately 3 miles) on Wightman Fork is the Summitville Mine Superfund Site. The Alamosa River flows in and out of Terrace reservoir before eventually flowing into the Rio Grande River south-east of the Town of Alamosa. The Rio Grande River flows southward out of Colorado into New Mexico. Terrace Reservoir is used as a source for irrigation as is the Alamosa River. The San Luis Valley is a productive agricultural area for the State of Colorado. The Alamosa River is an important part of the local agricultural economies. A map of the study area is shown in Figure 2.
Figure 2. Map of Terrace Reservoir and vicinity
[4]
Final TMDL
Figure 3. Satellite Image of Terrace Reservoir
Water quality in Terrace Reservoir is determined by the water quality of the Alamosa River upstream from the reservoir. The Alamosa River water quality, in turn, is determined primarily by pollutant loadings from various tributaries draining hydrothermally altered areas. Much of the watershed is characterized by hydrothermally altered geology, and the tributaries carry elevated concentrations of metals and exhibit depressed pH levels. The tributaries and areas with this hydrothermally altered geology include the Stunner Altered Area, drained by Iron Creek, Alum Creek, Bitter Creek; the Jasper Altered Area, drained by Jasper Creek and Burnt Creek; and the Summitville Altered Area that lies in the upper Wightman Fork drainage. Hydrothermal alteration describes a process by which hot, mineralized fluids seep into [5]
Final TMDL
faults and fractures resulting from previous volcanic activity. Such alteration creates base metal and precious metal-rich deposits within the altered areas. Because the geology of the Upper Alamosa basin includes extensive areas of hydrothermally altered terrains, the region has an extensive history of mineral extraction. The Colorado Geologic Survey (“CGS”) published an inventory of mine adits and associated features located in the Upper Alamosa Watershed in 1995. The inventory included evaluations of over 300 mine adits and features. Three “moderately-sized” mines were identified within the watershed. The inventory also reported the existence of “dozens of naturally occurring acidic, metal-rich springs (“NOAMS”). Pollutant contributions from these mines are summarized in Kirkham et al., 1995. The dominant mines, in terms of Fe are, the Pass-Me-By Mine from the Stunner Altered Area and the Miser Mine and Guadaloupe Mine from the Jasper Altered Area. The most significant mine in the basin, both in terms of production history and water quality impacts, is the Summitville Mine, located in the upper portion of the Wightman Fork drainage. Placer gold was discovered in Wightman Fork in 1870. These placer deposits were reportedly worked until at least 1888. More significantly, in 1875 three mills, including the Summit Mine were put into operation. A fourth mine, the Reynolds adit, was completed in 1897. By 1915 the four operations were consolidated with the formation of Summitville Gold Mines, Inc. A series of ownership changes followed during which time additional exploration work was undertaken. Active mining was primarily conducted underground. Galactic Resources Limited acquired the property and initiated open pit/heap leach operations in 1984. The property was worked until 1992, when Galactic declared bankruptcy and abandoned the site. Galactic experienced problems managing wastewater discharges and runoff from the site soon after its operations commenced. Among the problems encountered were leaks from the underdrain system beneath the heap leach pad, acidic groundwater beneath the Cropsy waste dump and insufficient containment for runoff from waste rock piles. The lack of adequate containment resulted in surface discharges of untreated acidic, metal bearing wastewater. These releases had a catastrophic effect on aquatic populations, essentially depleting the aquatic community to a point below Terrace Reservoir. The U. S. Environmental Protection Agency (“EPA”) provided emergency response assistance upon Galactic’s declaration of bankruptcy. Emergency efforts focused on containment of impounded wastes from the heap leach operation. In May 1994 the site was placed on the National Priorities List of Superfund sites. TMDLs for several segments on the Alamosa River for several metals (CORGAL03a (Al, Cu, Pb, Zn), CORGAL03b (pH, Al, Cu, Zn), CORGAL03c (pH, Al, Cu, Zn), CORGAL03d (pH, Cu, Zn), CORGAL05 (pH), CORFAL08 (Cu), and CORGAL09 (Cu) were approved in September 2007. Although the Terrace Reservoir (CORGAL08) does not attain its assigned FeTrec standard, and the sources of Fe-Trec obviously are from the watershed segments upstream of the reservoir, those segments have much higher Fe-Trec standards (12,000 ug/L) than Terrace [6]
Final TMDL
Reservoir (1000 ug/L). The upstream segments attain their assigned Fe-Trec standards. The water quality in the Alamosa River downstream of Wightman Fork has improved in recent years due to remediation efforts at Summitville. However, iron concentrations continue to exceed the Terrace Reservoir water quality standard. Background In 1993, the WQCC adopted ambient standards for iron, based on historic and recent data which indicated the presence of naturally elevated levels of iron in segments of the Alamosa River from Alum Creek to just upstream from Terrace Reservoir. Under the Basic Standards, the Commission may adopt ambient standards only where the ambient conditions are naturallyoccurring or are the result of irreversible human impacts. The standard of Fe(ch)=12000 ug/L adopted for the mainstem of the Alamosa River upstream of Terrace was determined to be consistent with pre-mining concentrations. Although this reach of the Alamosa River has been resegmented several times, and now includes 4 segments, the ambient iron standard of 12000 ug/L has been retained. Because of the presence of naturally elevated levels of iron, and based on evidence heard by the Commission during various rulemaking proceedings, the Commission was not convinced that the TVS of 1000 ug/l for iron was attainable in the river. However, the Commission did adopt TVS of 1000 ug/L for iron for Terrace Reservoir. Terrace Reservoir is designated Use-Protected because it was identified in the 1988 Section 305(b) Report as being impacted by a combination of metals loading and fluctuating reservoir levels. IV. WATER QUALITY STANDARDS Standards Framework Waterbodies in Colorado are divided into discrete units or “segments”. The Colorado Basic Standards and Methodologies for Surface Water, Regulation 31 (WQCC 2010), discusses segmentation of waterbodies in terms of several broad considerations: 31.6(4)(b)…Segments may constitute a specified stretch of a river mainstem, a specific tributary, a specific lake or reservoir, or a generally defined grouping of waters within the basin (e.g. a specific mainstem segment and all tributaries flowing into that mainstem segment.) (c) Segments shall generally be delineated according to the points at which the use, physical characteristics or water quality characteristics of a watersource are determined to change significantly enough to require a change in use classifications and/or water quality standards. As noted in paragraph 31.6(4)(c), the use or uses of surface waters are an important consideration with respect to segmentation. In Colorado, there are four categories of beneficial use which are recognized. These include Aquatic Life Use, Recreational Use, Agricultural Use, and Water Supply Use. A segment may be designated for any or all of the “Use Classifications”. 31.6 Waters shall be classified for the present beneficial uses of the water or the beneficial uses that may be reasonably expected in the future for which the water is suitable in [7]
Final TMDL
its’ present condition or the beneficial uses for which it is to become suitable as a goal. Each assigned use is associated with a series of specific numeric standards. These pollutants may vary and are relevant to a given Classified Use. Numeric standards are identified in sections 21.11 and 31.16 of the Basic Standards and Methodologies for Surface Water (WQCC, 2012).
[8]
Final TMDL
Uses and Standards addressed in this TMDL The Colorado Basic Standards and Methodologies for Surface Water, Regulation 31 identifies standards applicable to all surface water statewide (WQCC, 2012). The pollutant of concern for this TMDL assessment is Fe-Trec. Fe-Trec concentrations exceed the Aquatic Life Use-based standard intended to protect against long-term, sub-lethal (chronic) effects. The FeTrec standard for CORGAL08, Terrace Reservoir is a chronic standard of 1000 ug/L. There is no acute standard for Fe-Trec assigned to Terrace Reservoir. The specific numeric standards assigned to the Terrace Reservoir Segment 8 are contained in the Classifications and Numeric Standards for the Rio Grande River Basin, Regulation 36 (WQCC, 2012) (Table 3). Chronic and acute standards are designed to protect against different ecological effects of pollutants (long term exposure to relatively lower pollutant concentrations vs. short term exposure to relatively higher pollutant concentrations). Where chronic standards are assigned, they were used because they represent a more conservative approach than the acute standards. Chronic standards represent the level of pollutants that protect 95 percent of the genera from chronic toxic effects of metals. Per Regulation 31, chronic toxic effects include but are not limited to demonstrable abnormalities and adverse effects on survival, growth, or reproduction (WQCC 2012).
Water Quality Criteria for Impaired Designated Uses WBID
Impaired Designated Use
Applicable Water Quality Criteria and Status
CORGAL08
Aquatic Life Cold 2
Total Recoverable Iron / Not Attained
Applicable State or Federal Regulations: (1) Classifications and Numeric Standards for the Rio Grande River Basin, Regulation No. 36 Table 3. Ambient water quality criteria and status for 303(d) listed Terrace Reservoir.
The relevant standard for the reservoir segment addressed in this document is a Table Value Standard. Although many for metals are hardness-based and expressed as the dissolved fraction. The Aquatic Life Use-based TVS for iron is an exception and based on the total recoverable fraction and is not hardness based. The other uses assigned to Terrace Reservoir, Agriculture and Recreation, do not include iron standards. Therefore, the Aquatic Life Usebased standard for Fe-Trec is the only iron standard assigned to Terrace Reservoir.
V. PROBLEM IDENTIFICATION 303d Listing History [9]
Final TMDL
Terrace Reservoir was previously listed on the 303(d) list for impaired waters. It was included on the 1998 303(d) list for pH, copper, manganese, and zinc. Terrace Reservoir appeared on both the 2002 and 2004 303(d) lists for copper. TMDLs for these metals were approved in 2005. This TMDL addresses the Fe-Trec impairment for Terrace reservoir from the 2008, 2010, and 2012 303(d) lists. Pollutant Contribution To a great extent, iron loading in Terrace Reservoir has been the combination of natural geologic conditions and historic mining activities as opposed to permitted source discharges. Upstream of Terrace Reservoir, many of the tributaries of the Alamosa River flow through highly mineralized areas. Some of the names of the tributaries attest to the mineralized nature of the area with the names like Alum, Bitter, Silver and Iron Creeks. The segments of the Alamosa River, CORGAL03a, CORGAL03b, CORGAL03c, and CORGAL03d, which extend from above the confluence with Alum Creek to Terrace Reservoir, have been assigned Fe-Trec standards of 12,000 ug/L. These segments meet the assigned Fe-Trec standards, yet the load carried into Terrace Reservoir results in exceedance of the reservoir Fe-Trec standard. Terrace reservoir receives high loads of iron, and exceeds the assigned chronic standard of 1000 ug/L for Fe-Trec. Approximately 17 miles upstream of Terrace Reservoir is the Summitville historical mining site. This mining area is a superfund clean-up site that is listed on the Environmental Protection Agency’s National Priorities List. Although the Summitville Altered Area and associated Galactic mine site has been considered, before remediation, the greatest source of pollution to the Alamosa River, iron concentrations in the Alamosa River segments upstream from this influence are high enough to prevent attainment of the iron standards in Terrace Reservoir. Historically, environmental impacts from the Summitville site were mainly related to metals such as copper. A TMDL for copper was completed in 2005 for Terrace Reservoir. The Alamosa River and some of its tributaries flow downstream from the Summitville Superfund Site and into Terrace Reservoir. There is a water treatment facility on the Summitville Superfund Site that treats for metals. In 1994 and 1995, the USGS assessed metal transport through Terrace Reservoir (USGS, 1996). Concentrations of iron were higher in the Alamosa River upstream of the reservoir than downstream of the reservoir. The reservoir functioned as a sink for total iron, as well as other metals. During the study, approximately 75% of the total iron load transported into Terrace was retained in the reservoir. The highest daily loads of iron transported into Terrace Reservoir were observed during the snowmelt period, and the highest in-reservoir concentrations occurred from June through November. During the pre-peak and peak snowmelt periods, the iron loads transported into the reservoir were predominantly in the suspended fraction (>90%). However, during the post-peak snowmelt and summer-flow periods, approximately 60% of the iron load was in the dissolved fraction. Metals loads varied substantially in relation to stream flow and [10]
Final TMDL
concentrations. In the mid-1990s, Kirkham et al. studied sources of metals in the Alamosa River Basin outside of the Summitville Mining Area. They gave a conservative estimate that abandoned mines could contribute 11% of dissolved iron in the Alamosa River above its confluence with Wightman Fork. Total metal loadings would be even less. VI. WATER QUALITY GOALS Segment 08, Terrace Reservoir, is classified for Aquatic Life Cold 2, Recreation E, and Agriculture uses. The goal of this TMDL is attainment of the table value standard for Fe-Trec of 1000 ug/L in Terrace Reservoir. As stated previously, the Aquatic Life Use is the only assigned use on Terrace Reservoir that includes an iron standard. As such, this TMDL is written to the most stringent iron standard assigned to Terrace Reservoir.
VII. INSTREAM CONDITIONS Hydrology The hydrograph of the Alamosa River is typical of other high mountain streams. Low flows occur in the late fall to early spring, followed by large increases in flow during the annual snowmelt period. The snowmelt period starts in late spring, peaks in May or June, and tails off through the summer. Smaller tributaries show the same pattern, but tend to show greater influences from summer rain events. The period of record of the flow data used in this TMDL was from 1988 through 2008, although a shorter period of record was used in the figures to illustrate detailed hydrologic patterns. Figure 4 illustrates the hydrograph for the Alamosa River upstream from Terrace Reservoir. Figure 5 summarizes the interannual variability by month in the stream discharge upstream of the reservoir.
[11]
Final TMDL
Figure 4. Hydrograph of the Alamosa River upstream from Terrace Reservoir.
Figure 5. Box and whisker plots for median monthly flow on the Alamosa River above Terrace Reservoir. Boxes represent upper and lower quartile values while whiskers represent 5th and 95th percentile values. Asterisks indicate median flows. (Period of Record: Oct. 1988 thru Sep. 2008) In addition to review of stream flows of the Alamosa River upstream from Terrace Reservoir, reservoir water storage patterns were examined. Figures 6 illustrates annual patterns for reservoir storage from 2000 through 2008. This figure clearly illustrates that patterns of storage are mostly consistent over time. The amount of water stored is driven by water rights and is dependent on the amount of water available in any given year. [12]
Final TMDL
Figure 6. Terrace Reservoir Storage. The regular patterns in the hydrologic loading to Terrace Reservoir are emphasized because the hydrologic budget is an essential underlying component for estimating pollutant loads to the reservoir. The inflow quantity to the reservoir is dependent on availability of water based on climate conditions. The maximum storage in the reservoir peaks during April and May. Reservoir storage is lowest during August through October. Climate Climate data for the area is collected at the Del Norte weather station located near Terrace Reservoir and will be used for this TMDL. Climate data for this station was collected from January 1893 through June 2012. A summary of the climate data is provided below: Average annual precipitation (in.): 9.40 Month of highest precipitation: August (1.70 in.) Month of lowest precipitation: February (0.33 in.) Average annual snowfall (in.): 39.7 Average annual temperature: 42.8°F Month of lowest average temperature: January (5.7°F) Month of highest average temperature: July (78.7°F) (source: www.wrcc.dri.edu )
[13]
Final TMDL
VIII. ANALYSIS OF POLLUTANT SOURCES Ambient Water Quality Data Water quality data used for development of this TMDL include samples collected from within Terrace Reservoir as well as from the Alamosa River upstream of Terrace Reservoir. The reservoir data are used to evaluate the ambient condition of the impaired waterbody and to examine temporal and spatial patterns. Data for river samples upstream from the reservoir are used to link the reservoir to watershed loads delivered to the reservoir. The Terrace Reservoir samples were collected by USGS, RMC, and WQCD from 1994 through 2011. Alamosa River samples were collected by the WQCD, USGS, and RMC Consultants from 1991 through 2011 (n=35). Reservoir Data Fe-Trec concentrations in Terrace Reservoir samples are highly variable and frequently exceed the water quality standard of 1000 ug/L (Figure 7).
Figure 7. Terrace Reservoir Fe-Trec sample event means. The ambient condition of the reservoir was determined in a manner consistent with the Colorado Water Quality Control Division (Division) assessment methodology. The Division assesses Fe-Trec as the 50th percentile, or median of the available data. The 303(d) List assessment that identified Terrace Reservoir as impaired used reservoir data with a period of record from 2003 through 2006. That assessment calculated the 50th percentile, or median, of sampling event means. Table 4 presents the results of the assessment from the 2008, 2010, and [14]
Final TMDL
2012 303(d) Lists. For the development of this TMDL, additional data have been compiled. Data collected since 2006 are included for a longer period of record. An assessment of these data is included in Table 4 (2003-2011). Terrace Reservoir CORGAL08
Ambient Fe-Trec (ug/L) 1681.5
Fe-Trec TVS (ug/L) 1000
n=
POR
8
CORGAL08
1372
1000
17
20032006 20032011
Table 4. 303(d) Assessment of Terrace Reservoir Fe-Trec. Monthly median concentrations for Fe-Trec in Terrace Reservoir illustrate temporal characteristics in the reservoir. Sampling events after 2000 are conducted only during May-June and September-October. In order to examine other months, pre-2000 data were added to this part of the assessment and the monthly median values are presented in Table 5. Exceedances were observed in the months of May, June, and July. The median observed Fe-Trec concentration for September approaches the standard. No samples were collected during the months of January, February, and November. Ambient Fe-Trec ug/L (50th percentile) N/A N/A
Month TVS Fe-Trec ug/L January 1000 February 1000 March 1000 604 April 1000 143 May 1000 2313 June 1000 1300 July 1000 1223 August 1000 889 September 1000 977 October 1000 489 November 1000 N/A December 1000 465 Table 5. Ambient Water Quality for Terrace Reservoir (N=38)
Individual reservoir samples were used to evaluate spatial characteristics of Fe-Trec in Terrace Reservoir. Vertical profile samples were collected at 3 levels in the reservoir: top, middle, and bottom. Vertical variations in Fe-Trec concentrations were observed when the [15]
Final TMDL
reservoir is thermally stratified, typically with higher concentrations in the bottom layer (Figure 8). When the reservoir is mixed, the Fe-Trec concentrations from the different depths are fairly uniform. Note that the POR for this figure is 1995-2011. The data for 1994 were not included because information for sample layers was not available. When the individual layers are assessed for attainment of the total recoverable Fe standard, the bottom layer exceeds the standard, while the middle and top layer do not (Table 6).
Figure 8. Terrace Reservoir Fe-Trec, 1995-2011.
[16]
Final TMDL
Reservoir Layer Top Middle Bottom Table 6.
TVS Fe-Trec ug/L 1000 1000 1000
Ambient Fe-Trec ug/L (50th percentile) 715 810 1360
Box and whisker plots illustrate the monthly variability of total recoverable Fe in Terrace Reservoir (Figure 9). The boxes represent the 25th and 75th percentiles, while the whiskers represent the 5th and 95th percentiles. Medians are shown as blue dashes. The water quality standard for total recoverable Fe of 1000 ug/L is exceeded during the months of May through July.
Figure 9. Monthly Fe-Trec concentrations in Terrace Reservoir. Boxes represent the upper and lower quartile values while whiskers represent 5th and 95th percentile values. The blue dashes indicate median Fe-trec concentrations. (Period of record: 1994 thru 2011) Studies of Terrace Reservoir conducted by the USGS in the 1990s (USGS, 1997) concluded that inflow chemistry was the dominant controlling factor of metal chemistry in Terrace Reservoir. The reservoir is thermally stratified from approximately mid-May through August. During stratification, river underflow is predominant, meaning that due to temperature and density gradients, inflowing river water follows the lower level of the reservoir, and the FeTrec concentrations in the lower layer reflect the inflow concentrations. Although transport and deposition of suspended solids (including iron) varied spatially and temporally, most of the suspended solids were deposited in the reservoir. The lower layer of the reservoir remained oxic [17]
Final TMDL
during the stratified period so it is unlikely that iron was released from the sediments. This study concluded that 75% of the total iron that entered the reservoir remained in the reservoir, indicating that the reservoir was a sink.
Watershed Data: The source of water to Terrace Reservoir is the Alamosa River upstream from Terrace Reservoir, segment CORGAL03d. Data for this segment, immediately upstream of the reservoir are examined to estimate total recoverable iron loads to Terrace Reservoir. Data available for this analysis were for samples collected from the Alamosa River in segment CORGAL03d, upstream from Terrace Reservoir by the WQCD, USGS, and RMC Consultants from 1991 through 2010 (n=148). Monthly median total recoverable Fe values are presented in Table 7 to characterize the inflow water quality to Terrace Reservoir. Discharge values are monthly medians. These data were used to calculate median monthly loads delivered to Terrace Reservoir from the Alamosa River.
Month
Fe-Trec (ug/L
January February Mar April May June July August September October November December
3070 4832 6234 6632 5140 5665 3340 6600 4100 2850 3623 3070
Discharge (cfs) 16 16 20 70.5 345 289 75 51 38 31.5 20 16
Table 7. Median Total Recoverable Fe for Alamosa River upstream from Terrace Reservoir.
Data from several sites in the watershed are used to characterize the watershed pollutant sources to the Alamosa River and support development of load and wasteload allocations for the TMDL. These are discussed in the TMDL Allocation section.
[18]
Final TMDL
IX. TMDL ALLOCATION The Terrace Reservoir TMDL was developed by estimating the monthly in-reservoir FeTrec, and comparing these loads to an allowable load, which is the load that would be observed if the Fe-Trec standard of 1000 ug/L was attained. The required percent reductions are applied to the instream load delivered to Terrace Reservoir from the Alamosa River. Total Maximum Daily Loads A TMDL is comprised of the sum of the Waste Load Allocation (WLA), Load Allocation (LA), and a Margin of Safety (MOS). The Waste Load Allocation consists of the permitted and non-permitted point sources within the segment. The Load Allocation is the pollutant load attributed to natural background, also referred to as non-point sources. TMDL = Sum of Waste Load Allocations + Sum of Load Allocations + Margin of Safety Waste Load Allocation No permitted Waste Loads were identified upstream from Terrace Reservoir. The primary non-permitted sources from mining include the Summitville Mine, the Pass-Me-ByMine on Iron Creek, and the Miser Mine and Guadalupe Mine in Jasper Creek. In addition, hundreds of mining features have been identified in the watershed, however these are cumulatively minor contributors to the overall Fe loads carried in the watershed.
Load Allocation Sources considered non-point sources and not attributable to mining are identified under the Load Allocation of the TMDL. The dominant sources of total recoverable Fe to Terrace Reservoir are from natural background and fall under the Load Allocation. Margin of Safety In accordance with the Federal Clean Water Act, a margin of safety was included in this TMDL. The margin of safety is required to be included to account for the uncertainty about the relationship between the pollutant and the receiving waterbody. The MOS can be implicit or explicit. The Terrace Reservoir TMDL includes a 10% explicit MOS.
[19]
Final TMDL
Reservoir Loads In developing the total maximum daily load for Terrace Reservoir, in-reservoir observed median daily loads by month were compared to a monthly in-reservoir allowable load. The total recoverable iron data and median reservoir volume data discussed in previous sections were used to calculate observed daily iron loads by month for the reservoir. The allowable load was calculated using the reservoir standard of 1000 ug/L total recoverable iron and the median reservoir volume, and represents the in-reservoir load that would attain water quality standards. The resulting allowable load was used to calculate the percent load reductions that would be required for the reservoir to attain water quality standards (Table 8). Month
January February March April May June July August September October November December
Fe-Trec, observed (ug/L) 534* 534* 604 143 2313 1300 1223 889 977 489 477* 465
Fe-Trec, TVS (ug/L) 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000
Reservoir Volume (AF) 183 242 273 287 315 309 186 117 94 90 114 168
Observed Load (lbs/D) 266 352 448 111 1985 1095 620 284 249 120 148 213
Allowable Load (lbs/D) 498 659 742 781 858 842 507 320 255 246 309 458
% Reduction
No reduction required No reduction required No reduction required No reduction required
57 23 18 No reduction required No reduction required No reduction required No reduction required No reduction required
Table 8. Terrace Reservoir Fe-Trec: observed loads, allowable loads and required percent reductions. Bold type indicates months when the Fe-Trec standard is exceeded and when load reductions are required. (*no samples were collected during January, February, and November. For TMDL development, the average of adjacent months was used.)
Watershed Loads In general, the highest watershed loads were observed during the months that the Fe-Trec standards exceedances were observed in the reservoir. Fe-Trec load reductions are required in the reservoir during the months of May, June, and July. The TMDL was developed by applying the reservoir percent load reductions to the watershed loads delivered from the Alamosa River upstream from Terrace Reservoir. A ten percent margin of safety was applied to the TMDL (Table 9).
[20]
Final TMDL Month
January February Mar April May June July August September October November December
Fe-Trec, observed (ug/L) 3070 4832 6234 6632 5140 5665 3340 6600 4100 2850 3623 3070
Discharge (cfs) 16 16 20 70.5 345 289 75 51 38 31.5 20 16
Observed Load (lbs/D) 265 417 673 2525 9576 8840 1353 1818 841 485 391 289
Reduction (%)
TMDL Load (lbs/D)
10% MOS (lbs/D)
TMDL w/10% MOS (lbs/D)
4118 6807 1109
412 681 111
3706 6126 998
No reduction required No reduction required No reduction required No reduction required
57 23 18 No reduction required No reduction required No reduction required No reduction required No reduction required
Table 9. Total Recoverable Fe TMDL for Terrace Reservoir. Loads represent Alamosa River upstream from Terrace Reservoir. Bold type indicates months when Terrace Reservoir exceeds total recoverable iron standard.
Allocation Methodology A number of natural and anthropogenic pollutant sources occur in the Alamosa River watershed. As discussed earlier in previous sections, much of the upper basin lies within areas of hydrothermally altered terrain. Contributions of iron from these areas represent a significant source of loading. Naturally occurring acidic, metal-rich springs (NOAMS) have been identified as non-manmade sources. While it is difficult to separate contributions from man-made sources from the natural sources, long-term monitoring data and literature sources provide some guidance. Long-term monitoring data were used to estimate relative contributions of loads at different locations along the Alamosa River. Loads are presented as percentages. Although monitoring data were not available for all months, the results give reasonable estimates for drawing conclusions and determining allocations between wasteloads (man-made sources) and loads (natural sources). Monitoring data were used to calculate the relative proportion of loads coming from different parts of the watershed. Total recoverable iron loads from the Summitville Mine Water Treatment Plant effluent (SMWTP) are compared to the loads at the mouth of Wightman Fork. Loads at the mouth of Wightman Fork are compared to Alamosa River loads upstream of Terrace Reservoir. These relative contributions are presented in Table 10. The loads from SMWTP are a small percentage of the loads at the mouth of Wightman Fork. During the spring, when Fe-Trec exceedances are observed in Terrace Reservoir, SMWTP makes up less than 2% of the load from Wightman Fork. Later in the summer, the relative contribution of SMWTP to Wightman Fork is somewhat higher, but still less than 10%. The relative contribution of SMWTP to Terrace Reservoir would be even lower. In addition to the treatment plant, relic [21]
Final TMDL
mining features may also contribute to the Fe load in Wightman Fork. However, this is unquantified. The relative contribution of Fe-Trec loads from Wightman Fork to the Alamosa River upstream from Terrace Reservoir also is presented in Table 10. Wightman Fork contributes 3% to 4% of the Fe-Trec load during the months that exceedances in Terrace Reservoir are observed. The relative contributions of Fe-Trec loads from the Alamosa River upstream from Wightman Fork are shown in Table 10. Loads from this part of the basin represent over 50% of the load to Terrace. During September and October, the relative contributions were greater than 100%. This suggests that some of the Fe-Trec load may settle out between the comparison sites. Studies by the U.S.G.S. noted this as well. The remainder of the loads may be attributed to other tributaries in the watershed. Alamosa River upstream of Wightman Fork as % of Alamosa River upstream of Terrace Reservoir.
Summitville WTP effluent as % of Wightman Fork
Wightman Fork as % of Alamosa River upstream Terrace Reservoir
April
1.22
6.3
57
May
0.44
3.0
61
June
0.30
2.3
Month January February March
July
0.0
August
4.05
92
September
4.10
4.3
139
October
7.94
7.0
117
0.0
November December
Table 10. Relative contributions of select sources of Fe-Trec loads to Terrace Reservoir. Not all sources are represented.
Additional information for developing allocations is provided in Kirkham et al., 1995. Conservative estimates from this report suggest that draining mines could be responsible for nearly 11% of the iron in the Alamosa River upstream from the confluence with Wightman Fork. A significant portion of this iron can be attributed to the Pass-Me-By Mine. Ultimately, however, the majority of iron carried in the watershed originates from natural sources. [22]
Final TMDL
For the purpose of allocating the observed Fe load for Terrace Reservoir between waste load and load sources, it is assumed that the conservative estimates from Kirkham et al. are applicable for the entire basin. The observed river load is partitioned into an observed wasteload (11%) and observed load (or non-point sources) (89%). For the TMDL, the allowable load less the Margin of Safety is assigned to the Load Allocation. The Wasteload Allocation will be allocated 0 lbs/D (Table 11).
Month
January February Mar April May June July August September October November December
Observed Load (lbs/D)
265 417 673 2525 9576 8840 1353 1818 841 485 391 289
Point source contribution of Observed Load
29 46 74 278 1053 972 149 200 93 53 43 32
Nonpoint source contribution of Observed Load 236 372 599 2247 8522 7868 1204 1618 749 431 348 257
Reduction (%)
TMDL Load (lbs/D)
10% MOS (lbs/D)
No reduction required No reduction required No reduction required No reduction required
57 23 18
4118 6807 1109
412 681 111
No reduction required No reduction required No reduction required No reduction required No reduction required
Table 11. TMDL Allocations for Alamosa River Total Recoverable Fe.
[23]
Waste Load Allocation (lbs/D)
0 0 0 0 0 0 0 0 0 0 0 0
Load Allocation (lbs/D)
3706 6126 998
Final TMDL
X. RESTORATION PLANNING AND IMPLEMENTATION PROCESS Water Quality Improvements in the Watershed Improvements to the watershed primarily are the under the Summitville Superfund Program. Additional remediation projects may be implemented under the Alamosa River Watershed Restoration Master Plan. However, at this time, no funding sources have been identified.
Monitoring Monitoring under the Superfund Program is ongoing.
XI. Public Involvement The public has had the opportunity to be involved in the Water Quality Control Commission (WQCC) hearings since 1996 when the WQCC adopted ambient-based standards. Opportunities have also been available through the 303(d) listing process which also has a public notice period for public involvement. This TMDL was available for public comment during the month of August 2012. The Division received one set of comments from the public notice. These comments and the Division’s responses are in Section XIII. The Division submitted a final TMDL report and response to comments to EPA for approval in October 2012. After reviewing the comments, and based on stakeholder comments sent directly to EPA after the Public Notice, EPA requested that the Division include a 10% Margin of Safety. EPA also requested that the Margin of Safety be expressed as part of the inflow load from the Alamosa River, rather than as part of the inreservoir load. The Division made the requested changes to the TMDL report.
[24]
Final TMDL
XII. REFERENCES Ferguson, Sheryl, and Edelmann, Patrick, 1996, Assessment of metal transport into and out of Terrace Reservoir, Conejos County, Colorado, April 1994 through March 1995: U.S. Geological Survey Water Resources Investigations Report 96-4151. Kirkham, R. M. and Lovekin, J. R. 1995. USFS Abandoned Mine Land Inventory Project Summary Report, Rio Grande National Forest—Conejos Peak Ranger District. Colorado Geological Survey. Kirkham, R. M. Lovekin, J. R., and Sares, M. A., 1995, Sources of acidity and heavy metals in the Alamosa River Basin Outside of the Summitville Mining Area, Colorado. Proceedings: Summitville Forum ’95.Colorado Geological Survey Special Publication 38. Stogner, Sr., Robert W, Edelmann, Patrick, and Walton-Day, Katherine, 1997, Physical and Chemical Characteristics of Terrace Reservoir, Conejos County, Colorado, May 1994 Through May 1995: U.S. Geological Survey Water Resources Investigations Report 96-4150. WQCC 2012: Colorado Department of Public Health and Environment, Water Quality Control Commission, Section 303(d) List Water-Quality-Limited Segments Requiring TMDLS, Regulation #93, Adopted February 2012. WQCC 2008: Colorado Department of Public Health and Environment, Water Quality Control Commission, Classification and Numeric Standards for Rio Grande Basin, Regulation No. 36, Effective January 1, 2012. WQCC 2012: Colorado Department of Public Health and Environment, Water Quality Control Commission, The Basic Standards And Methodologies For Surface Water, Regulation No. 31, Effective September 30, 2012.
[25]
Final TMDL
XIII. RESPONSES TO COMMENTS RECEIVED CONCERNING THE AUGUST 2012 PUBLIC NOTICE DRAFT The Terrace Reservoir TMDL report was Public Notice during August 2012. The Division received one set of comments on the TMDL. The comments were from the Law Office of John M. Barth on behalf of the Alamosa Riverkeeper. The comments and the Division’s responses are below.
Alamosa Riverkeeper : The Barth letter describes perceived problems with the TMDL stating: The WQCD is using old, incomplete, and unrepresentative dataIn calculating the TMDL, the WQCD uses old, incomplete, and unrepresentative data. For example, the WQCD largely relies on data from a 1995 Kirkham report. See, pages 22 and 23 of the Draft TMDL. The WQCD largely relies on data from a 1995 Kirkham report (pg 22-23). These data are 17 years old and fails to represent current conditions. Summitville WTP came on-line in 2012 and will dramatically alter the water quality of Wightman fork, the Alamosa River, and Terrace Reservoir. Alamosa Riverkeeper believes that the TMDL is premature because it does not take into account the water quality due to the new treatment plant. Alamosa Riverkeeper requests that WQCD delay promulgation of a TMDL for Terrace Reservoir and instead immediately implement a three-year comprehensive water quality monitoring plan in the watershed in the Upper Alamosa River in order to update the data it relies on for this TMDL and better understand the changes to water quality resulting from the operation of the Summitville Water Treatment Plant.
Division Response: The TMDL report used several sources of data. The TMDL was developed using recent monitoring data (as described in Section VIII Analysis of Pollutant Sources, Ambient Water Quality Data, page 14). The period of record for these data extends from 1991 through 2011. The 1995 Kirkham reports were used to provide information about relative contributions from sources in the upper watershed. The Division found no evidence that would suggest conditions in these areas have changed significantly. No development has occurred in these areas. [26]
Final TMDL
Although a new water treatment plant came online in 2012, the TMDL used monitoring data from the treatment plant that was active since 2000. The period of record for this was 20002011. Effluent data from the old treatment plant indicate discharge concentrations were consistently less than 1000 ug/L for total recoverable iron. In developing the TMDL, the Division also examined data from Wightman Fork at the mouth (POR=2000-2011). The median value of the total recoverable iron data at this site was 910 ug/L, less than the Terrace Reservoir standard. The Division assumes that the new plant will operate as well, or better than the old plant. Because, as discussed above, there is adequate data to characterize Terrace Reservoir water quality through 2011, the Division sees no need to delay the TMDL. Alamosa Riverkeeper: The data relied upon by the WQCD is also incomplete. TMDL admits that discharges from mining point sources in Wightman Fork are “unquantified.” Moreover, the TMDL admits, the primary non-permitted sources from mining include the Summitville Mine, the Pass-Me-By Mine on Iron Creek, and the Miser Mine and Guadalupe Mine in Jasper Creek. The TMDL then goes on to assume that these point sources contribute little iron contribution to Terrace Reservoir, but fails to provide current factual support for its claim. The TMDL then ignores all point source contribution in the watershed by failing to assign any wasteload allocation to these point sources (pg 23)
Division Response: The Division examined and assessed the total recoverable iron concentrations from the Summitville Water Treatment Plant. The TMDL includes the quantification of the relative contributions from the treatment plant effluent discharge. The TMDL states on page 21-22: “In addition to the treatment plant, relic mining features may also contribute to the Fe load in Wightman Fork. However, this is unquantified.” Ultimately, however, as stated in the Division’s response in the previous section, the monitoring data for Wightman Fork at the mouth indicate that the total recoverable iron concentrations are below the Terrace reservoir standard of 1000 ug/L. If data for a segment is in attainment of relevant standards, it is not necessary to quantify individual WLAs.
[27]
Final TMDL
Based on information from the Kirkham reports, the Division estimated the relative contributions of the non-permitted mining sources. As stated in the Division’s earlier response, there is no evidence suggesting significant changes to Alamosa Basin upstream from Wightman Fork. Therefore, it is appropriate to base these estimates on information from the Kirkham reports. Furthermore, relative contributions from Wightman Fork and from the Alamosa River upstream from Wightman Fork were based on recent monitoring data (See Division response above discussing data period of record).
The TMDL does not ignore point source contributions. The TMDL effectively allows no wasteload contributions. The entire allowable load is allocated to the Load Allocation. However, the statement on page 23 is not clear, although Table 11 is correct. The last statement on page 23 should state “For this TMDL, Wasteloads will be assigned allocations of 0 lbs/D.” The text in the TMDL report will be modified to clarify this concept.
Alamosa Riverkeeper: The Barth letter goes on to list excerpts from Kirkham and Lovekin, 1995: In Iron Creek effluent from the Pass-Me-By mine portal imparts significant loadings of iron and aluminum. The headwaters of the Alamosa River in the segment between Iron Creek and Wightman Fork are severely degraded by iron, aluminum, and to a lesser extent manganese which are in part attributable to the Pass-Me-By mine. Wightman Fork [in which the Summitville mine is located] appears to contribute the majority of the copper and zinc loadings and also adds considerably to the manganese, iron and aluminum loading. Barth states that while the 1995 Kirkham reports provide a qualitative assessment of the relative contribution of point and non-point source pollution in the watershed, the data is old, and unrepresentative, and incomplete. Such data does not provide the recent representative data needed to conduct an accurate TMDL.
Division Response: Following the excerpts above, Kirkham and Lovekin, 1995, concluded that “Remediation of the drainage from the Pass-Me-By mine could perhaps result in significant improvement to the iron and aluminum loadings in Iron Creek and also perhaps to the Alamosa River in the section immediately below its confluence with Iron Creek, but it would have little or no effect on the [28]
Final TMDL
river below its confluence with Wightman Fork and perhaps not even in the segment below the inflow from Alum Creek.” While the Division did not have recent data for the upper reaches of the Basin discussed by Kirkham and Lovekin, the Division did assess recent monitoring data from Wightman Fork and from the Alamosa River above the confluence with Wightman Fork. Assessments of recent monitoring data from Wightman Fork demonstrated that Wightman Fork total recoverable iron does not exceed the Terrace Reservoir standard of 1000 ug/L. As stated in Division Responses above, there is no evidence to suggest that changes have occurred in the upper reaches of the Alamosa Basin upstream from Wightman Fork that would influence water quality. Therefore, it is appropriate to apply the information provided in the report to estimate relative contributions from the point and nonpoint sources. The Division applied the Kirkham report’s estimates of relative contributions of mining sources to the recent monitoring data to estimate mining and nonpoint source loads to Terrace Reservoir. These estimates were presented in the TMDL report.
Alamosa Riverkeeper: The Kirkham report referenced in the TMDL also identifies numerous significant mining point sources in the watershed that may be impacting Terrace Reservoir. Additional excerpts from Kirkham et al., 1995: “219 mine opening and 130 mine dumps in the watershed, all of which are “point sources” under the Clean Water Act; “an adit adjacent to Forest Road 250 north of Terrace Reservoir” that may be impacting water quality in the reservoir; “water was draining out of or standing within thirty-one of the inventoried mine openings” “Acidic water seeped from the toe of three of the min dumps…” “Mine drainage locally contained significant concentrations of iron [140 mg/l] and aluminum…” “Drainage from the Pass-Me-By mine was the most acidic…” “Dissolved water samples were collected during our investigation from Iron Creek above and below the drainage from the Pass-Me-By mine. Dissolved iron raised from 0.88 to 2.7 mg/l…” The Pass-Me By mine “could account for up to 21 to 31% of the dissolved iron in the creek at its mouth” with the Alamosa River. [29]
Final TMDL
“the abandon mines could be responsible for nearly 11% of the iron” in the Alamosa River.
Division Response: The excerpts above are incomplete, and it is not clear what point the commenter is trying to make. However, Kirkham concluded “Remediation of the drainage from the Pass-Me-By mine could perhaps result in significant improvements to the iron and aluminum loading in Iron Creek and also perhaps to the Alamosa River in the section immediately below its confluence with Iron Creek, but it would have little or no effect on the river below its confluence with Wightman Fork and perhaps not even in the segment below the inflow from Alum Creek.” And “A conservative estimate of the maximum possible contribution of mining to the degradation of the Alamosa River above the confluence with Wightman Fork can be made if it is assumed that all dissolved metals associated with draining mines eventually reaches the river and stays in solution. Under this scenario the abandoned mines could be responsible for nearly 11% of the iron and almost 18% of the aluminum, but only around 1% of the copper, manganese, and zinc in the river above the confluence with Wightman Fork at the time of sampling, with the balance of these dissolved constituents being attributable to natural degradation.“
And “A worst case estimate of the relative contribution of the Pass-Me-By mine drainage upon the Alamosa River can be developed by comparing the loads in the mine drainage to the loads in the mouths of Iron, Alum, and Bitter Creeks. Under such a scenario, the dissolved iron loading in the Pass-Me-By mine drainage amounts to about 3.1 to 4.3% of the combined dissolved iron from the three tributaries. For dissolved aluminum, the Pass-Me-By mine drainage comprises around 2.8 to 3.3% of the combined loadings from the three tributaries. Dissolved manganese, copper, and zinc loadings from the mine amount to only 0.2%, 0.4 to 0.7% and 0.6 to 1.5%, respectively, of the combined loadings from the three tributaries. A similar comparison using total metal loadings results in even lower percentages.” And A similar comparison between the dissolved metal loadings in the Pass-Me-By mine drainage and the Alamosa River above Wightman Fork showed the mine drainage was approximately 11 % of the Alamosa River above Wightman Fork. Kirkham and Lovekin (1995) concluded that if
[30]
Final TMDL
the Pass-Me-By mine drainage was eliminated as a source of metals, the loads in the Alamosa River could at best be improved only by this amount.
Alamosa Riverkeeper: Moreover, the Summitville mine site was excluded from the Kirkham Report and thus provides little guidance on the current impact of the mine site on the Alamosa River or Terrace Reservoir.
Division Response: The Division assessed recent data from Wightman Fork at the mouth and the Summitville Water Treatment Plant effluent. Total recoverable iron concentrations for both sites were below the Terrace Reservoir standard of 1000 ug/L.
Alamosa Riverkeeper: Finally, the Kirkham reports contain water quality sample results for dissolved iron, not total recoverable iron.
Division Response: Conclusions in the Kirkham reports suggested that for total metals, the relative contributions of mining vs natural sources would be even lower than those observed in the dissolved metals. However the Division used the percentages identified for dissolved metals, which the Kirkham reports concluded represented the maximum possible contribution of the mining sources and therefore was a highly conservative estimate. This conservative approach provides an implicit margin of safety.
Alamosa Riverkeeper: The TMDL fails to set any daily pollution limits from point sources
Division Response: The TMDL includes Wasteload of 0 lbs/D for point sources. The entire allowable load is allocated to the Load Allocation (background and nonpoint sources).
[31]
Final TMDL
Alamosa Riverkeeper: The margin of safety is vague and unverifiable
Division Response: TMDLs may contain an implicit or explicit Margin of Safety. In the Terrace Reservoir, conservative assumptions were used that may overestimate the required loading reductions. These include applying the in-lake percent reduction to the loads delivered from the Alamosa River upstream of Terrace Reservoir. The TMDL allocations also were based on the conservative relative contributions of mines identified by Kirkham for dissolved iron, which was 11%. The Kirkham reports suggested that this percentage would be lower for total metals. The Division decided that applying an additional 10% MOS would be overly conservative in this case. Subsequent to the Division’s final submittal for TMDL approval, and in response to receiving additional comments from the Riverkeeper’s attorney, EPA directed the Division to include an additional 10% MOS and to express the MOS as part of the inflow load from the Alamosa River. These changes were made to the TMDL.
Alamosa Riverkeeper: The TMDL fails to include an implementation plan
Division Response: Development or inclusion of an implementation plan is not a required element of the TMDL. However, the Alamosa River Watershed Restoration Master Plan, Alamosa River Foundation et al., July 2005, describes activities to address remediation of point and nonpoint sources within the Alamosa watershed.
Alamosa Riverkeeper: A monitoring plan should be developed and implemented.
Division Response: A monitoring plan is not a required element of the TMDL. In preparation for review of the basin-specific water quality standards, the Division conducts more intensive monitoring. Review of Regulation 36, Classifications and Numeric Standards for Rio Grande Basin (5 CCR 1002-36), takes place on a five-year schedule. In addition, the Colorado Department of Public Health and Environment Hazardous Materials and Waste Management Division contracts with TetraTech Routine to maintain a Summitville database related to the Summitville Superfund Site. Under this contract, TetraTech conducts water quality monitoring in the basin, collecting high flow and low flow water quality samples [32]
Final TMDL
annually from sites on Wightman Fork, the Alamosa River, Terrace Reservoir, and select tributaries.
[33]
