Test Pumping Playbook

Beschreibung

Groundwater bore test pumping hydraulic analysis
Nathan Littlewood
Flussdiagramm von Nathan Littlewood, aktualisiert more than 1 year ago
Nathan Littlewood
Erstellt von Nathan Littlewood vor mehr als 4 Jahre
1
0

Zusammenfassung der Ressource

Flussdiagrammknoten

  • Collate field notes
  • Cleanse data and apply corrections (baro, temp)
  • Graph: elapsed time, displacement, flow rate, activity
  • Plot hydrograph and annotate
  • Establish likely conceptual flow model
  • Download and view logger data
  • Obtain all required bore and aquifer input data
  • Check units and conversions
  • Review published T & S
  • Set up file for analytic model
  • Use software wizard and review tutorials
  • Plot data and perform Diagnotics
  • Check derivative curve catalog
  • Review log-log and semi-log 
  • Flat derivative indicates IARF and validity of C-J straight line
  • Does flow model fit with geology?
  • Choose appropriate solution
  • Compare with other solutions and expected values
  • Perform sensitivity analysis on T & S, Kv/Kh
  • Are analysis assumptions valid?
  • Use C-J semi-log straight line on late-time data when IARF
  • Check Aqtesolve course notes
  • Check for d-d curve inflections
  • Residual d-d late time is near graph origin (reverse plot)
  • Logan (confined) T=1.22.Qmax/smax
  • Logan (unconfined) T=2.43.Qmax.bsat/(smax.(2b-s))
  • Unconfined aquifer S=Sy (0.1-0.3)
  • Specific Capacity=Q/s (m3/d/m)
  • Skin Factor Sw= Aquifer Loss: BQ  Well Loss: CQ
  • Batu-Driscoll (unconfined) T=1.042(Qmax/smax)
  • Batu-Driscoll (confined)  T=1.385(Qmax/smax)
  • Screen over water table: Semilog plot inflection. Fit later data
  • Early straight line unit slope of both plots on log/log = well bore storage dominant
  • If derivative does not plateau then don't use straight-line analysis
  • Early Data: bore storage Mid Data: flow model ID Late Data: boundaries
  • Model-building procedure: 1.data mgt 2.diagnostics 3.model selection 4.curve matching 5.statistical analysis
  • Derivative smoothing
  • Deviation from unit slope as more water comes from aquifer
  • log-log derivative drops to zero if leaky aquifer or constant head boundary
  • The tail of derivative curve can be unreliable
  • C-J gives best results for single well tests as removes skin effect
  • Single well S biased by skin. Very low S = redevelop bore!
  • Partial penetration gives similar curve to skin
  • Theis recovery residual d-d method: t/t' intercept close to 1 = infinite aquifer. <1=no flow boundary. >1=recharge boundary
  • Recovery methods:  Argawal time Theis curve match Argawal C-J plateau late time match Pap-Coop match d-d & recovery data Residual d-d Theis late time thru origin
  • Perform approximation calculations
  • Compare multiple values of T&S
Zusammenfassung anzeigen Zusammenfassung ausblenden

ähnlicher Inhalt

Business Englisch
cathyc
Proteinbiosynthese
Cornelius Ges
Nathan der Weise Quiz
Laura Overhoff
Öffentliches Recht I -Teil 1
kicker03
Wortschatz Französisch 3. Gesundheit und Medizin
l_u_n_a_19
WERB Univie (mögliche Fragen)
frau planlos
PuKW STEP6 - Hummel (Sofort überprüfbar)
Tim Schröder
Vetie Radio 2014
S J
GESKO A JOUR WS 2019/20
Caroline Hannah
Repro 2016 Vetie
Julia To
Vetie - Recht 2020
Birte Schulz