CE 394k GIS in water resources

FINAL PROJECT REPORT

Beatriz Garcia-Fresca

12/5/02

POTENTIAL RECHARGE TO GROUNDWATER FROM MAINS LEAKAGE: AUSTIN, TEXAS (USA)

1. Introduction

2. Data collection and treatment

3. Spatial analysis and calculations

    leakage

    relationship with faulting

4. Results and discussion

5. Conclusions

6. Future work

7. References

 

Appendix: Building a Hydrologic Network

    elevation

    flow direction

    tracing

    discussion

1. INTRODUCTION

It is broadly recognized that recharge of rainfall to groundwater is inhibited in urban areas due to the presence of impervious cover. The present study discusses the fact that urban development can also produce the opposite effect, i.e. the increase of groundwater recharge. Water and wastewater mains leak, and lawns and parks are often over-irrigated, resulting in a potential increase on recharge to the groundwater, from strictly urban sources (e.g. Lerner, 1986, and Foster, 1996). Increases in recharge have been identified in cities all around the world, as shown in figure 1.

This study covers the Austin metropolitan area, which sits on the inactive Balcones fault zone, in central Texas (USA). The city occupies most of Travis County, and small parts of the surrounding counties.

Figure 2: location of the study area

ArcGIS is used to determine the leakage form the water distribution network to the Austin subsurface. The water distribution network will be superimposed to the geology of the Austin area in order to assess the potential recharge from the mains to the different hydrogeological units. The City of Austin recognizes a water loss in the distribution lines of 12% of the delivered water as shown in the newspaper clip in figure 3 (Austin American-Statesman, 1998).

Figure 3: unaccounted for water in Texas

The potential for the leakage to recharge to the groundwater depends to some extent on the geologic materials underlying the mains, as well as faulting.

2. DATA COLLECTION AND TREATMENT

Initially a large and varied amount of datasets were collected (see Future work section), the ones used in this study are the following:

• Geology

Source: UT BEG (Trembley, Andrews and Dale, after Garner and Young, 1976)

Description: lithologic contacts and faults of the Austin Area

Figure 4: Geology dataset

Click here to see the footprint of the city over the geologic map

• Water distribution network

Source: City of Austin GIS datasets

Description: line and point shapefiles of the distribution lines. Attributes include year installed, diameter, material, etc.

Figure 5: water distribution network dataset

• Jurisdictions

Source: City of Austin GIS datasets

Description: polygon shapefiles of the outlines of the different city jurisdictions. Only the “full purpose” jurisdiction was used.

Figure 6: Austin jurisdiction limits dataset

The principal data processing requirement consisted of representing all datasets in the same projection. The NAD 86 State Plane Texas Central (US ft) coordinate system was chosen, which was designed for this particular area and therefore minimizes distortion.

Minor processing consisted on assigning the standard colors to each of the lithologies displayed in the geologic map (figure 4).

3. SPATIAL ANALYSIS AND CALCULATIONS

Recharge to the groundwater depends, in part, on the permeability of the outcropping materials. For this reason, the first step is to reclassify the geologic formations into permeability units, by editing the attribute table of the shapefile containing the geology. The permeabilities assigned are listed in table 1, and correspond to the qualitative descriptions of the rocks by Duffin (1991), and Brune & Duffin (1983). The spatial distribution of the permeability units is shown in figure 7.

Table 1: characteristics of the rocks outcropping in the study area

Figure 7: relative permeabilities of the geologic materials in the study area

Click here to see the footprint of the city over the permeability map.

The Edwards Group, Georgetown Formation, and the alluvium are the potentially most permeable materials. The former two form the Edwards Aquifer. The Glen Rose Limestone may not be as locally permeable as the Edwards and Georgetown limestones, but it contributes its drainage to the recharge of the Edwards Aquifer. River deposits can also present important permeability values. According to this interpretation, three different hydrogeologic environments can be distinguished:

 Cretaceous Karstic limestones: very permeable locally, and susceptible to pollution. The outcrop of the aquifer is shown in figure 8.

Figure 8: outcrop of the Edwards aquifer in the Austin area

Low permeability materials don’t allow the direct infiltration of water to the underground but could contribute drainage to adjacent aquifer units, and baseflow to creeks.

Quaternary deposits overly both limestone and impermeable materials (figure 9). In the study area they consist on fluvial alluvium and terrace deposits, which can reach significant permeability values. They are often closely related to surface waters, and can present perched water tables.

Figure 9: Quaternary deposits in the Austin area

leakage

This section describes the calculations performed on ArcGIS to estimate the potential leakage from mains.

The water loss for the city of Austin is 12% of the water pumped into the distribution system. According to Thornton (2002), about 60% of the losses are “real losses”, and the remaining fraction “apparent losses”. The real losses correspond to the water lost to pipe leakage and breaks, and doesn’t include metering unbalances, thefts, etc. According to this, this study will adopt the value of 7% real losses, or a “leakage coefficient” (ℓ) of 0.07.

Water usage data from the City of Austin Water and Wastewater Service was used as the input to the calculations. Table 2 shows the historical water demand for Austin, for the period between 1993 and 2001.

Table 2: historical water demand, Austin, Texas

Source: City of Austin Water and Wastewater Utility

The input total annual pumpage chosen was that of 2001, i.e. 50,140 millions of gallons, which is an average of 364,442 m³/day. The volume of the mains the product of the diameter and length, both attributes of the dataset. The average daily average pumpage of water is then divided by the total volume of mains, and multiplied by the volume of each pipe, to obtain the daily volume of water delivered by each pipe. This value is then multiplied by the leakage coefficient to obtain the volumetric leakage per pipe. These calculations can be summarized by the following expression:

L = (x_i · D_i · P_TA · ℓ) / (S [x_i · D_i] · 365)                                                                          (1)

where,

x_i – length of each pipe

D_i – diameter of each pipe

P_TA – total annual pumpage

ℓ – leakage coefficient

A useful figure is the leakage per unit length of pipe (LL_i), defined as:

LL_i = L/x_i                                                                                                                               (2)

Such simple calculations could be comfortably carried out in a spreadsheet, except for the fact that these particular datasets have too many records for the capacity of the Microsoft Excel spreadsheet.

relationship with faulting

Austin sits on the inactive Balcones Fault Zone, which is oriented NE-SW (see figure 4). Faults represent potential preferential pathways for water to reach the subsurface. Water leaking immediately above a conductive fracture is likely to recharge the aquifers below. ArcGIS allows the superposition of pipes and faults (figure 10), and the identification of the lines crossing above fractures (figure 11). Then the volume of water leaking above fractures, per unit length of pipe is computed.

Figure 10: geometric relationship between distribution pipes and faults

Figure 11: distribution pipes crossing over fault traces

4. RESULTS AND DISCUSSION

Table 3 summarizes the calculations and results of this study. The real losses are determined as 7% of the 520,631.78 m3/d pumped into the distribution network. The 36,434 m³ which leak on average every day, represent a potential groundwater recharge of 21.94 mm per year, over the area covered by the city.

Table 3: summary of results