0003 Attachment 2 - Geotech and Environmental Assessment Bldg 100.pdf

Spec.Vol. II of V

Proj. No. 463-500 ~ Department of ~Veterans Affairs

Spee if i cations --------------

For:

At:

New VA Outpatient Clinic & Regional Office

VAMCAlaska Anchorage, Alaska

Issue March 23, 2007 rAmendmenf'

Open Bids May 3, 2007

Property of Department of Veterans Affairs

WITHIN 10 DAYS AFTER DATE OF OPENING BIDS, RETURN

THIS SPECIFICATION TOGETHER WITH DRAWINGS, POSTAGE

PREPAID TO:

Livingston Slone, Inc 3900 Arctic Blvd, Suite 301 Anchorage, AK 99503

No. Date

PART 1 - GENERAL:

1.1 DESCRIPTION:

ATTACHMENT "B"

GEOTECHNICAL DATA

This section covers geotechnical data and studies.

1. 2 APPLICABLE PUBLICATIONS:

A. Kranich & Associates Consulting Engineers "Potential for Groundwater Source for Cooling at VA Clinic" dated July 23, 2006.

B. DOWL Engineers "Subsurface Exploration Veteran's Affairs Outpatient Clinic

Elmendorf Air Force Base, Alaska" dated January 2006.

PART 2 - PRODUCTS:

Not used.

PART 3 - EXECUTION:

Not used.

--- END ---

00330 - 1

THIS PAGE HAS BEEN INTENTIONALLY LEFT BLANK

00330 - 2

Kranich & Associates Consulting Engineers

9780 CARLSON RD., ANCHORAGE, AK 99507

PHONE (907) 346-1901 FAX (907) 346-1169 July 23, 2006

{!o1...1---------------- I I ;.

i

I I i

I I I I ,ft -r~ l ffi

I I

TO:

RE:

Whomever It May Concern Potential for Groundwater Source for Cooling at VA Clinic

FROM: Robert W. Kranich, P.E.

INTRODUCTION:

The VA Clinic slated for construction south of the existing Composite Medical Facility at Elmendorf AFB, Alaska will require a source of energy for cooling purposes. The question was asked if the existing groundwater cooling system at the hospital has sufficient excess capacity to provide the cooling needs of the Clinic as well. I understand that the projected peak demand of the Clinic would be 500 ± gpm.

RESEARCH:

· 1. I reviewed available reports from the DOWL archives concerned with the development of the existing groundwater cooling system. The most useful document WdS a "Letter Report" by Golder Associates dated 3/7/97. (Exhibit B).

That report provides a thorough analysis of the capabilities of the hospital groundwater system as it was constructed.

2. I interviewed Bill Friars, the Assistant Project Manager for BSA/LB & B; the operation and maintenance contractor who operates the hospital central plant systems.

3. I attempted to locate and enter test well TW-3 located in the southwest comer of the hospital tract. That test well would have provided the best indication of long term effects of the existing system on the groundwater resource in the area. I located the well, but found that it had been decommissioned - cut off 4' below grade, filled up and capped with concrete.

FINDINGS:

1. I was given a tabulation of total gallons pumped and peak gpm for the months of May through July of 2005. (Exhibit A). Observations from that data were:

A. The lowest average flow rates represented by the data set were approximately 250 gpm. Average minimum cooling •tnon-summer" demand used for design was 400 gpm. (See Exhibit A, page 12).

B. The highest peak demand - nearly 800 gpm - occurred in the last few days of June. Maximum Anchorage temperatures on those days per National Weather Service were in the mid to high 701s. (Exhibit C). (Maximum recorded Anchorage temperatures are in.the mid 80's.)

C~ Mr. Friars confirmed that all systems in the hospital were operational during that time, so the data may be considered ~ typical.

2. Hospital plant maintenance does not keep record of well water levels, or individual pump production rates. All that they could tell me at the time of our interview - 7 /11/06 - was that at that time #3 well was pumping at 490 gpm with a water level of 181' and had been on lead duty for about 24 hours.

At that time water levels in Wells#.! & 2 were 124' and 126' respectively.

3. Well pumps in the existing wells were selected to develop the full production potential of each well. Increasing the pump capacity as has been suggested would not increase the capacity of the system.

ANALYSIS:

The dearth of data available prevents a detailed analysis of the state of the groundwater resource in the vicinity of the hospital. The data I was able to collect, however, did allow me to reach several general conclusions. They are:

1. The data set from May through July of2005 (Exhibit A) indicates that the "average non-summer demand" of the system is in the order of250 gpm or less.

The system design assumed that component of demand to be approximately 400 gpm (Exhibit B, page 12).

2. The highest pumping rate shown by the data set was 786 gpm which occurred on 6/28/05. On that day the high Anchorage temperature was 78°. (Exhibit C).

Assuming that space cooling at the hospital begins at outside air temperature of 55°, (per Mr. Friars) and allowing for a non-summer base demand of250 gpm, a straight-line projection of peak demand to estimate the maximum pumping rate required on a day of maximum Anchorage temperature of 86° would be:

786 gpm - 250 gpm/78° - 55° = 232 gpm/degree (86° - 55°) X 232 + 250 = 969 gpm

That rate compares well with the 1200 gpm projection of the original design when we consider that the base demand appears to be some 150± gpm lower than the design anticipated.

3. The static or non-pumping water levels on 7/11/06 at Wells #1 & 2 of 124' and 126' respectively, indicate that the water level in the aquifer has been drawn down 20' to 25' from the static levels of 100'± recorded during the construction period. This drawdown is on the order of what should be expected from the ongoing usage.

CONCLUSIONS:

As a result of my investigation, I have reached the following conclusions:

1. The overall demand on the system is somewhat lower - probably on the order of 10-15%-- than the design projections anticipated.

2. Well performance is still adequate to fulfill peak system demands with one well remaining in reserve.

3. Aquifer water levels have not been seriously drawn down.

4. Although it does not appear that the existing demands at the hospital are taxing the system, insufficient data exists to predict how much additional capacity would be available.

RECOMMENDATIONS:

Since there appears to be the possibility of sufficient groundwater available for the ._ cooling of the prOf!OSed clinic, I would suggest:

1. Assuming that aquifer conditions are suitable, construct another production well on the Clinic site and connect it to augment the hospital groundwater system. .

2. Connect the Clinic cooling system to receive it's requirements from the hospital system.

3. Include in the design of the Clinic's systems provision for future installation of mechanical chillers which may become necessary to cover peak cooling demands.

This approach would have the following advantages:

1. An additional production well would improve the total capacity and reliability of the hospitaJ/clinic groundwater cooling system.

2. With an additional well, the system should have sufficient capacity to meet all cooling demands of both facilities.

3. If the groundwater system were in future years found to become unable to accommodate peak cooling demands of both facilities, mechanical chillers could be installed at the clinic to assist.

· In any case, a monitoring program should be established at the hospital to test on an annual basis the specific capacity of each production well so that declining well efficiency can be detected before it causes serious difficulties. Also, the existing monitoring system should be programmed to provide continuous record of each wells production, well water levels, and outside air temperature. That data could be presented· in graphical form and archiveq ·on digital storage devices. It would allow· analyses of aquifer capabilities and enable the defense of water rights in the future.

Robert W. Kranich, P.E.

Enclosures: Exhibits A, B & C

EXHIBIT A

Elmendorl WelfWater Usage MoS-d--'3/. tff,f,, /1_,,./, IJ .,,,,, /, /}.,p ~

MTH DAY TOTAL~! PEAKGPM_,, MTH DAY TOTALGPM' PEAKGPM V MTH DAY TOTAL~ PEAKGPM

5 5 352713 I 253 .t~4- 6 1 402167- 346 7 1 851622 732 ,5''1 I 5 6 398895 .. _ 337 2.77 6 2 407045 338 7 2 764887 625 5 7 402630 259 'Z.,£30 6 3 426854 378 7 3 730721 603 5 8 396856 325 L76 6 4 441799 383 7 4 673822 574 5 9 403124 364 '2.Sr:? 6 5 496879 384 7 5 599928 601 5 10 390631 333 2. 71 6 6 487746 433 7 6 618972 667 5 11 351508 389 2.fir 6 7 448879 369 7 7 704466 650 5 12 434323 353 30/ 6 8 436928 396 7 8 6'55415 612 5 13 404638 370 Z.21/ 6 9 482039 427 7 9 654124 598 5 14 401674 474 -'/?'I 6 10 509056 487 7 10 525441 489 5 15 346264 281 ~ r 6 11 519914 477 7 11 551508 476 5 16 367271 324 6 12 489278 429 7 12 534323 435 5 17 386134 335 6 13 488471 445 7 13 584638 501 5 18 396569 349 6 14 555528 528 386 7 14 571674 484 5 19 422922 402 6 15 610782 572 ~24- 7 15 546179 381 5 20 410421 350 6 16 813466 625 S-tS- 7 16 567273 401 5 21 384603 304 6 17 651015 611 4-5".Z. 7 17 486134 335 5 22 393582 334 6 18 654193 578 1/--ff- 7 18 496569 355 5 23 426341 436 6 19 425~31 413 7 19 422922 394 5 24 390223 361 6 20 437771 386 7 20 410421 350 5 25 387937 331 6 21 532042 519 7 21 384603 344 5 26 412122 356 6 22 496630 398 7 22 393582 334 5 27 392847 328 6 23 500989 439 7 23 406341 427 5 28 378892 329 6 24 511285 445 7 24 394423 351 5 29 407257 391 6 25 595854 533 7 25 387937 323 5 30 484110 350 a 26 594020 679 7 26 381122 365 5 31 412857 367 6 27 730183 741 .S--0 7 7 27 392853 328

10737344 6 28 822530 786 i,"7/ 7 28 378900 321

6 29 864644 735 /4 OP 7 29 315267 378

6 30 959250 776 Lb/,, 7 30 344335 332

16792868 7 31 327781 343 16058183

EXHIBITB

Letter Report

Modification P00036

Additional Test Wells

CMF, 3rd Medical Center

Eln1endorf AFB, Alaska

March 12, 1997

Contract No. DACA85-92-C-000 I

A- 0 O O .1 ·

RESULTS OF THE PW-2 PUMPING TEST:

ELMENDORF AFB HOSPITAL

1.0 INTRODUCTION

This d?cument summarizes the results of the step test and constant-rate pumping test of the newly mstalled production well (PW-2) at the Elmendorf AFB hospital site. See Well Log Attachment D. In addition, a brief overview of ~he studies conducted to date is included fo;

background purposes, and an updated estimate of future wellfield performance is provided.

Attachments A, B, and C contain figures, plots, and tables showing the results of the step test, constant-rate test, and wellfield performance evaluation, respectively.

2.0 SUMMARY OF PREVIOUS INVESTIGATIONS

The work conducted to develop a groundwater supply for the hospital can be grouped in three phases. Phase I involved the drilling and testing of test wells TW-1 and TW-2, and the installation and testing of production well PW-I . Phase II involved the drilling and testing of test wells TW-3 and TW-4, and conducting a long-term pumping test of PW-1. Phase III involved the installation and testing of PW-2 at the site of TW-4. The results of the first two phases are briefly summarized below. The results of Phase III are presented in the final sections of this document. .

Phase I: TW-1, TW-2, and PW-1

From the work conducted under this phase, the characteristics of the aquifer had been deterinined to a limited degreet and preliminary estimates of wellfield yield were developed. The aquifer transmissivity and storativity was estimated within a certain range (transmissivity: 2,400 fl:2/d to 2,900 ft2/d, storativity: 3 x 104 to 5 x 10-4).

However, aquifer characteristics that could potentially have a major influence on well and wellfield yield, such as leakage and boundary effects were unknown at the time.

Based on the limited knowledge of the aquifer characteristics available, the number of wells required to meet the groundwater supply demand was estimated at between 3 and 5 wells. ·

Phase II: TW-3, TW-4, and Long-Term Testing ofPW-1:

From the work conducted under this phase, the characteristics of the aquifer were determined to a much greater degree, and the potential wellfield performance was better wtderstood. Estimates of aquifer transmissivity and storativity were refined (Transmissivity: 2,800 ft.2/d, Storativity: 2.5 x 1 o·5), the results of which indicated that the aquifer transmissivity was . towards the upper portion of the range established earlier, and that storativity was significantly less than previously estimated. From the long-term pumping test, leakage was interpreted to be minimal, and the aquifer was determined to be laterally extensive. Leakage of water from an overlying or underlying aquifer into the target aquifer would have enhanced the potential yield of the aquifer. In contrast, if the aquifer had proven to be limited laterally, potential aquifer yield would have been less. Based on the limited leakage, the aquifer appeared to be somewhat less capable of meeting groundwater system demands than previously estimated.

A key finding of the Phase II investigation is that the aquifer is heterogeneous, with pockets of higher permeability gravel embedded in a less pe~eable aquifer. The

A 0002

March 7, 1997 - 2 - 963-5289.007 presence of pockets of high permeability gravel will afford greater well performance of wells installed in the gravel pockets. This, however, adds an additional degree of complexity to the groundwater exploration program in that, the only way to determine whether or not a gravel pocket is present is by installing and testing a highly efficient test well. Installation of a monitoring well and monitoring drawdown in response to pumping of a different well will not reveal whether or not the monitoring well is located in a higher than average permeability zone. The test well drilling and testing program undertaken for this project was, therefore, fortuitous in this regard. Other commonly used groundwater exploration programs would not have been capable of determining the heterogeneous character of the aquifer.

Although well performance can be enhanced if the well is installed in a gravel pocket, the interference effect, or the drawdown at other wells depends on the average aquifer characteristics. In other words, the heterogeneity of the aquifer can help with respect to individual well yield, but it does not reduce the negative effects of interference. In this particular case, interference effects are a critical issue with regard to the perfonnance of the wellfield.

Based on the updated understanding of the aquifer characteristics from the Phase II investigation, the number of wells required to meet the groundwater supply demand was refined as follows:

• 1 well (PW-2) to meet average 400 gpm demand and average summer demand (850 gpm);

• 2 to 3 wells (to meet peak daily demand of 1,200 gpm); and

• 3 to 4 wells (to meet fire flow demands of 1,500 gpm).

The above estimated wellfield design is based on a much better understanding of the aquifer. With the better understanding of the aquifer characteristics, it became clear that the aquifer is capable of supplying the total volume of groundwater necessary to meet total system demands. However, given constraints on the nwnber of wells and well spacings, the drawdown in each well may have to be maximized in order to meet design peak system demands.

The remainder of this document describes the work conducted under Phase III and an updated evaluation of wellfield performance.

3.0 PUMPING TEST OF pw ... 2

3.1 Field Work

3.1.1 Step Test

A stepped-rate pumping test was conducted in PW-2 on February 17, 1997. The test consisted of four one-hour steps conducted at rates of 268.0, 345.5, 437.0, and 488.0 gpm.

Water levels were monitored in the pumping well, and in observation wells TW-2 and TW-4 by a Golder Associates engineer. Total drawdown at the end of the final step was about 65 feet. The results indicated that the well was relatively inefficient, and that the well needed further development. M-W Drilling subsequently redeveloped_ the well, and a second step test was perfonned on February 21, 1997.

Golder Associates A 0003

March 7, 1997 - 3 - 963-5289.007

The second step test consisted of six one-hour steps conducted at rates of268.0, 345.5, 437.0, ~88.0, gpm! 655 gpm, and 802 gpm. \3/"ater levels were monitored in the pumping well, and m observation wells TW-2 and TW-4 by a Golder Associates engineer. Total drawdown at the end of the final step was about 55 feet, indicating that the additional development was successful in improving the well performance. A hydrograph of the step test is shown in Figure A-1. ·

3.1.2 Constant Rate Pumping Test

An 8-hour constant-rate pumping test was conduct~d on February 24, 1997. A pumping rate of 7 40 gpm was selected for the constant rate pumping test, based on the maximum capacity of the test pump, and the need to maintain a constant pumping rate as the water levels are drawn down in the well. Water levels were monitored in the pumping well, TW-2, TW-3, and TW-4. At the end of the test, the drawdown in the pumping well was ah<?ut 52 feet.

3.2 RESULTS

3.1 • Step Test Results

The step test results are depicted graphically in Figure A-1. As shown in the figure, drawdown at the end of the first step at a pumping rate of 268 gpm was about 12 feet.

Drawdown at the end of the sixth and final step at a pumping rate of 802 gpm was about 55 feet. The relatively small drawdown occurring during the step test demonstrates the high performance of the well. ·

The issue of well performance has been discussed in detail in previous reports. Due to its ·' importance with regard to the ability of the wellfield in meeting design demand requirements, however, the concepts are presented again here.

Well performance is typically viewed as the efficiency of getting water from the aquifer into the well. In highly efficient wells, the water levels in the wellbore are nearly the same as the water levels in the aquifer immediately adjacent to the well. In low efficiency wells, the water levels in the wellbore are substantially lower than the water levels in the aquifer immediately adjacent to the well as a result of head losses, or friction losses, as the water enters the well. These head losses are caused by turbulent flow of water as it passes through the screen into the well.

Figure A~2 shows the results of an analysis of well perfonnance using the conventional Jacob method (Jacob, 194 7). The Jacob analysis is the most widely used method for estimating well performance. Using the Jacob method, the amount of drawdown caused by turbulent groundwater flow and the amount of drawdown caused by linear groundwater flow can be determined independently. The method equates drawdown associated turbulent flow to "well loss" and drawdown associated with linear flow to "aquifer los~". Aquifer loss is the natural drawdown occurring in the aquifer in response to pumping due to the hydraulic properties of the aquifer (transmissivity and storativity).

The Jacob method provides good way to quantify the amount of turbulent flow losses at any given pumping rate through the calculation of a turbulent well loss coefficient (C). The amount of turbulent well losses can be calculated as: s=CQ2

, wheres is well losses in feet, C is the turbulent well loss coefficient in ft/ gpm2

, and Q is the pwnping rate in gpm. The value of C was calculated to be 3.29 x 10·5 ft/gpm2 for PW-2 using the Jacob method. The figure

Golder Associates A 0004

March 7. 1997 -4- 963-5289.007 shows ¢at the well losses attributed to turbulent flow are less than 21 feet for pumping rates of up to 800 gpm.

Knowing the value of the turbulent well loss coefficient, along with • aquifer hydraulic parameters, the drawdown in the pumping well can be estimated at any pumping rate at any given pumping duration. This approach was used in the fonnulation of the model used to predict the hospital wellfield perfonnance (Golder, 1995).

Figure A-3 shows the observed drawdown during the step test in comparison to the theoretical drawdown, given the average hydraulic properties of the aquifer and well loss characteristics (the value of C), as determined from the Jacob method discussed above. The observed drawdown is roughly half the theoretical dra'\iydown. This demonstrates the results expected at PW-2 based on the testing of TW-4 (the nearby test well). As discussed in the Ground Water Supply (QWS) report (1995) that summarized the results of the Phase II site investigation, the drawdown during testing of well TW-4 was less than the theoretical drawdown. The reason for the apparent discrepancy between the observed drawdown and the theoretical drawdown is that the aquifer is heterogeneous, and that the well was installed in a pocket of high penneability gravel. In order to account for the enhanced performance of the well, the concept of a "skin effect" was introduced, as presented in the GWS. Briefly, it is a way of accounting for conditions when the permeability of the aquifer immediately adjacent to the well screen differs from that of the aquifer as a whole.

The skin effect varies linearly as a function of the pumping rate, or ss = BQ, were S is the skin effect in feet (negative in this case), B is a skin effect coefficient in ft/gpm, and Q is pumping rate in gpm. The coefficient representing the skin effect can be determined by trial and error matching of the theoretical drawdown with the observed drawdown .

. Along with the theoretical drawdown assuming average aquifer conditions, Figure 3 also shows the previously estimated performance of PW-2 before it was installed (dark solid line), and the updated theoretical match ( dashed line).

3.3 · Constant Rate Pumping Test Results

Figure B-1 shows the drawdown that occurred in PW-2 during the constant-rate test conducted at a rate of 740 gpm. Drawdown during the first IO minutes of pumping was affected by wellbore storage and possibly pumping rate adjustments. After the first 10 minutes of pumping, the drawdown was about 3 5 feet. Between 10 minutes of pumping and the end of the test (480 minutes), water levels were drawn dQwn an additional 17 feet, which equates to a total drawdown of about 52 feet. The estimated aquifer transmissivity from the drawdown data is about 2,500 ft'/d, which is consistent with that measured at TW-4 and the other test wells.

Figures B-2, B-3, and B-4 show the drawdown observed in observation wells TW-4, TW-2, and TW-3, respectively. The drawdown at each of the wells at the end of the 8-hour test is also presented in Table 1. As sho~ in Table 1, the drawdown in the observation wells is significant. In PW-1, for example, the drawdown was about 17 feet or about 17% of the total available drawdown at this site (100 feet).

4. UPDATED ESTIMATE OF WELLFIELD YIELD

Several simulations were conducted using the spreadsheet wellfield model developed previously for the site (Golder, 1995) to evaluate the potential wellfield yield. The estimates are updated from those presented in the GWS report based on the known performance of PW-

Golder Associates A 0005

March 7, 1997 -·5 - 963-5289.007

2, (instead of the anticipated performance of PW-2), and the plans to install a replacement well near PW-1.

Model fonnulation and assumptions have been provided in previous reports. Some of the assumptions, including those that have been revised are provided in Section 3 .1. Results are presented in Section 3.2.

3.1 Assumptions

For modeling purposes, the following assumptions·were made:

• Aquifer transmissivity is 2,800 ft2/day;

• Aquifer storativity is 2.5 x 10·5;

• No leakage occurs;

• The aquifer is fully recharged on a yearly basis;

• The maximum available drawdown is 100 feet;

• Well loss coefficient is 2.2 x 104 ft/gpm2 for PW-1 (approximate performance);

• Well loss coefficient is 2.5 x 10·5.ft/gpm2 for PW-la{assumed);

• Well loss coefficient is 3.39 x 10·5 ft/gpm2 for PW-2 ( based on the well performance test);

• Negative skin for PW-la (TW-1 site) is 0.0055 gpm/ft (used to account for slightly more permeable conditions anticipated to exist at this locality, estimated · from pumping test ofTW-1); and

• Negative skin for PW-2 (TW-4 site) is 0.059 gpm/ft (used to account for more permeable conditions at this locality, based on the well performance test).

The performance of the yet-to-be-installed PW-la was estimated based on a review of the test results of TW-1. From this analysis, it is anticipated that the performance of well PW-1 a will be somewhat better than the average well installed in the aquifer, but will probably not be as good as PW-2.

As discussed in the GWS report, the model used to estimate future wellfield performance does not match the observed interference effects at the well sites exactly, in part, because average aquifer parameters must be assumed in the model. For example, Table 1 shows the predicted interference effect versus the .observed interference effect at PW-1 (TW-2 site) and TW-3 from the 8-hour pumping test of PW-2. This table shows that the interference effect at PW-1 was widerestimated by about 2.5 feet, and underestimated at TW-3 by about 0.6 feet.

Previous comparisons of predicted versus observed interference effects were presented in the GWS report (Figure D-1).

Due to the limitations of the model under heterogeneous conditions and other inherently uncertain conditions, such as long-term static water level conditions, the modeling results must be considered as approximations.

A 0006

March 7, 1997 963-5289.007

3. 2 Modeling Results

The predicted wellfield performance presented in Table 2 is very similar to that presented in the GWS report, largely because the performance of PW-2 was predicted accurately. Table 3 summarizes the results of a series of various wellfield configuration simulations, and the estimated drawdowns that would result. Attachment C includes a more detailed summary of the wellfield simulations summarized in Tables 2 and 3.

As summarized in Table 2, a single well, PW~2; will provide a large proportion of the water needed during normal system operation, and is c_apable of providing the entire volume of water needed to meet all system demands, if sufficient surface storage were available.

However, given the expense and operational difficulties associated with surface storage, the present design criteria is to be capable of meeting peak demands by pumping groundwater directly. Under this design_ criterion, an additional well (PW-la) and possibly a third well (PW-1) is needed to meet design peak daily chiIIing demands of 1,200 gpm, and a third well (PW-1) and possibly a fourth well may be needed to meet peak design system demands of 1,500 gpm for 2 hours.

The reason why one well is capable of providing almost all of the system demands, but that several wells are required to meet peak demands is due to interference effects between the pumping wells, under conditions when the available d.rawdowns in the wells are nearly maximized, as shown in Table 3. Interference effects reduce the net incremental wellfield capacity for each new well added to the wellfield. For example, the first well installed in a hypothetical aquifer may provide 800 gpm to the system at maximum drawdown, whereas the fourth well may contribute only an addition 200 or 300 gpm to the system. .

The uncertainty in the final number of wells needed to meet peak demand stems primarily ·· from the combination of the unavoidable limited accuracy at which drawdowns and interference can be predicted in response to pumping, and the need to maximize the drawdowns in each well to meet peak demands.

5.0 CONCLUSIONS

The following conclusions are drawn from the analysis conducted for this investigation:

• Well PW-2 is highly efficient, and capable of yielding a large portion of the total amount of groundwater needed. The ·high performance of the well stems from its installation within a pocket of higher permeability gravel and from limited well losses;

• The successful installation of PW-2 demonstrates that high efficiency wells can be installed in the aquifer;

• The heterogeneous nature of the aquifer has been confirmed;

• The predicted number of wells and wellfield performance is very similar to that presented in the GWS report: .a single well, PW-2 can provide a large proportion of the water needed during normal system operation, and a total of 3 wells (possibly 4 wells) required to meet design peak demands; ·

• In order to meet design peak demands for the 3-well system, the drawdown in the wells is predicted be at the design maximum (100 feet), which may not be advisable, without an adequate backup supply; and

Golder Associates A 0007

March 7, 1997 - 7 - 963-5289.007

• The need for additional wells is a result of interference drawdown, and the limited area available for installing the wells. Interference effects reduce the net incremental wellfield capacity for each new well added to the wellfi.eld.

6.0 RECOMMENDATIONS

Recommendations are as follows, under the assumption that the installation of an additional well (PW-1 a) will proceed as planned: - _ -

• Conduct a step test and 8-hour constant-rate test to determine the performance of Well PW .. la, and monitor the other wells to document interference effects; and

• Update estimates of wellfield performance, and determine _whether or not additional wells are needed to meet system demands.

6.0 REFERENCES

Anderson DeBartolo Pan, Inc. 1995, Ground Water Supply: Evaluation of Alternatives Composite Medical Facility 3rd Medical Center, Elmendorf AFB, Alaska.

Golder Associates, July 6, 1995. Revised Addendum to Water Well Data Analysis and Recommendations, Composite Medical Facility, Eln;tendorf AFB, Alaska. Prepared for The Department of the Army U.S. Army Engineer District, Alaska.

Jacob, C.E. 1947. Drawdown Test to Determine the Effective Radius of Artesian Well.

Trans. Amer. Soc. of Civil. Engrs., Vol. 112. Paper 2321.

Golder Associates A UOOB

TABLES ·

A 0009

3/10/97. 963-5289.007

TABLE 1

DRAWDOWNS AT THE END OF 8-HOUR PW-2 PUMPING TEST:

OBSERVED AND PREDICTED

Observati9n Wells

PW-1 TW-3 PW-2

Observed Drawdown (ft) 17.25 12.4 52.15 Predicted Drawdown (ft) 14.7 11.8 52.54

Difference (ft) 2.55 0.6 -0.39

* Using updated spreadsheet weUfield model (Golder, 1995)

A O O 1 o

C>

-... 4LE2

SUMMARY OF ESTIMATED WELLFIELD YIELDSC1)

'"400 gpm 850gpm 1,200e;pm 1,500 ra,m

· .365-Da:5, 79-Day 16-Hour 2-Hour Pumping Na.of Pumping Puinping Pumping Pumping Scenario Wells Wells Rate Rate Rate Rate

ALLWELL 1 PW-2 Yes Yes No No ALL WELL 2 PW-2" PW-la Yes Yes Yes.., No ALL WELL 3 PW-2,PW-la,PW-1 *,. • Yes Yes Yes Ya•• ALL WELL 3 PW-2,PW-la,PW-l*it.- Yes Yes Yes Yes•

,. Criteria of maximum allowable drawdown of 100 feet exceeded by less than 5 feet fur base ca&e assumptions.

** Criteria of maximum allowable drawdown of 100 feet exceeded by Jess than 15 feet for base case assumptions_ *0 Assumes existing poor effi.ciency of PW .. 1 (Efficiency ooefficient = 2.2 x 10 "'-4 It/gpm,,.., 2) ••** Assumeseffidencyof PW-1 is improved (Efficiency coefficient-- 2.5 x l_0.,... .. 5:ft/gpm"2)

Yes = Pumping rate and duration can be met based on assumptions included in the text No = Pumping rate and duration cannot be met based on assumptions included in the text

1) Model formulation and assumptions presented in previous reports.

CO,

I

I

Q

. l

· sffiY BY= sr:ATTLE

3(f/97

3- 7-87

TABLES

OOWL-ANCHORAGE;#l2

963--5289.007

SlMULA TED DRAWDOWNS FOR VARIOUS WELLFIELO SIMULATIONS case Simulated Cases Orawdown No. PW-1 PW•2 PW-1a

Average Non-Summer Demand 400 aom: 385 days {ft) (ft) (ft) 1 2Wells~ PW-2, PW-1a wttn best estimate of performance) 23-70 39.70 24.00 2 2weus, vc:led (PW..2, PW-1a with best estimate of performance) 23.70 39.70 24.00 3 2Wells 'JW-2, PW-1 a with same perfonnance as PW-2) 23.70 39.70 24.00 4 3Wells1 .1W-2. PW•1a with best estimate of performance. PW-1) 23.70 3.9.70 24.00 5 3Wells ::,w .. 2, PW-1a with best estimate of pertonnance, PW-1) 23.70 39.70 24.00

Average Summer Dav Demand ISO gpm: 79 days Drawdown 1 2 Welts CPW-2, PW.•1 a with best estimate of s;,ertormance) 46.10 92.75 46.70 2 2 Wells cveled (PW-2. FW-1 a wHh best estimate of oerformanoe} 59.30 62.40 88.43 3 2WeJls PW-2, PW-1 a with same oerrormance as PW-2) 59.30 62.4U 65.95 4 3We11s, PW-2, PW-1a with beSt estimate of oerrormanca. PW-1) 46.10 92.75 48.1C 5 3 Wells (PW-2. PW-1a with best estimate of performance. pw .. 1) 48.1 92.8 46.1

., Peak Summar ChiUing Demand 1,ZOO aDm: 18 hours Drawdown 1 2 Wells CPW-2, PW-1a with best estimate of Derformance} 64.90 101.55 86.98 2 2 Wells 1 :vcled CPW-2, PW-1 a with best estimate of oerformance) 87.80 93.55 97.23 3 2Wells PW-2. PW-1 a With same narfonnance as PW-2) 87.80 93.55 74.75 4 3 Wells PW•2, PW-1 a with best estimate of performance1 PW-1) 64.90 101.60 87.00 s 3Wetls PW-2, PW-1a with best estimate af pertonnancea PW-1) 64.90 101.60 87.00

Fire Demand after Peale Summer Chillina Demand 1500 oom: 2 hours Orawdown 1 2 Wells lPW-21 PW-1a with best estimate of performance) 77.40 105.55 122.73 2 2 Wells 1 ~eled (PW .. 2, PW•1a With best estimate of oertonnance) 75.90 109.60 117.30 3 2WeJls :aw.2. PW-1 a With same ner1onnance as PW-2) 80.30 97.55 95.95 4 swells ~-2, PW-1a with best estimate of oelfonnance, PW-1) 113.D0 10SAO 89.50 5 3 Wolls (PW-2. PW--1a With bast estimate of perfonnance, PW-1) 95.40 105AO 99.50

Total Avaifable Drawdown 100.0 100.0 100 Note: Balded drawdowns mdJCate VIOiation of 100 ft maximum available draWdown criterion.

A 0012

PW-2 STEP TEST RESULTS ATTACHMENT A

A O O 1 3

C)

3/7/97 eo g

I 30 i I [ ..

! I !

i I i L ••• I t i

( 287gpm .... ~ .. ,.

. .A.. I ~·

0 50

.j

100 150

Time (min) l 802 gpm

300 350

963::S289.007

J

• Observed Drawdown l

FIGURE A-1

PW-2 STEP TEST HY.DROGRAPH

c.v J co

____ ___j

3nJ97

0.07 .

t

C = 3.29E-5 f1/gpm"2· t

B = 3.49E•211/gpm I

0.05

0.05

Iumul~at Well Losses:

t ~ --------268 gpm: 2.4 ft i 488 gpm:7 .8 ft t ~ 802 gpm: 21.2 ft _,_

. i-

I ~ .

L----< -----0.04

D.03

0.02

0.01

<.n 0 100 200

•••••••• ., .. "•••· 00 ,•,,•.•--•••,J,, .......... •,•~•-:,: .•• 0 •••••··· .. ··••----•' .. ,_.,•,•, .. ,, .. •• ·•••• •·•• • ... -, ... ,oHO•••••po'••· ''"••·•••••ON••·••·· 0,o •• ·•••••• -

300 400 500 600

Pumping rate (gpm)

700 800 900

FlGUREA-2

JACOB WELL PERFORMANCE METHOD

co -.l l>

C)

C>

CJ\

3/7197

-e.

C:

'O lo;.

D

• Observed Drawdown .,,,----

-f-.--

- - - Theoretical Match

-- -Previously Estimated Performance

-Average Well Perfonnance ,,-~ I I i i ' 1

I ·( 2137gpm 1 i

J rl _.J r ~

0 50 100 l 802gpm J I I

-- .;:. .... ...,._ • : :,-t .r c ........ ....... ~► _., I l I f

I

Time (min)

300 350 400

FIGUREA-3

PW-2 STEP TEST HYDROGRAPH WITH ESTIMATED WELL PERFORMANCES

Cu I

---.J I to -.l

(J co O>

8 r;

Vl

VJ

-3 rrJ r-r er

I

~ n is Fri· "(;j: ,.., C.::

PW-2PUMPINGTESTRESULTS. ATTACHMENTB

A O O 1 7 e -

Pumping Well PW-2 Obs. Well PW-2

55...,.---,---,---,-----,-.--.----,----,--,--,---------,r----r---r-----r----.-------.--"T""T""~-----r---.--,---,-----,,-,-.,.............----.-------

Transmissivity = 2,500 ft A 2/d Storativity = NA so-- Radial Dist. from Pumping Well = O ft Pumping Rate= 740 gpm

X

X X

45-;---t---r-----t--r--t-t---H-t------ir---t---+--+-+-++-+·i------+--+---f-+-l-+-+-H------+-~+--+---I-J..-U-U

C: )<

~ 40-t---t------t---t---t--t--i--t-+-t------ir---+--+---t-----++--H---l------t--------1----...g:~ KL-1--+-11------l-H----~-+-----'--l-...J,......J.-t.....LI

1 -0 X

X 35-t---+---+---+-+-+---IH-t-+---t-----+--+---+---+-t---+-++---+--+--l-+-l-+---1-+-1-----1---'----l---l--J.._J_~~~ :K

)c

X 3Q--t---+----+---+-+--+--lf-H-t---l----+--+--+-++-+-+4------1---+--I---J--J--+--~---1--+-----l--l-----L-LJ.J._j

25-+---t----+---t--t--+-t---H-+-------if----+----+--+-++-H-+--, ---+--+--+-+---+--1-if--f-+----+--~~~--1-1-~

0.1 1 10 100 1000

I Jacob-Cooper Method I Time (min) Figure B•1

Golder Associate$

D C)

\0 g 22

C:

~ 20 '"C

Cl 18

:K

)K ~K

I Jacob-Cooper Method J

Pumping Well:PW-2 Obs. Well: TW-4

•I, :k

'It-

)K

-'1'

X 'I, X

;i,..

10 100 Time (min) x>

)k'.x 'V'

X

X

X

X

Figure 8•2