ALERT! ALERT! CORRECTION TO LAST LECTURE Suppose that source enters the up gradient end of a column At a continuous concentration of Co=1000mg/l K = 0.1 cm/sec dh = 10 cm dl = 100 cm φ = 0.2 02 Dispersivity αx = 5 cm What will the concentration be at 50 cm after 1000sec? cm sec 10cm = 0.05 cm average linear velocity 0.2 100cm sec cm distance traveled in 1000sec? d = v t = 0.05 1000 sec = 50cm sec Kdh v= = φdl 0 .1 By inspection we expect the concentration should be 0.5*Co=500mg/l But let’s carry out the calculation ALERT! ALERT! CORRECTION TO LAST LECTURE – EXCEL WAS CORRECT cm cm m2 10000cm2 cm2 v = 0.05 Dx = v α + D* = 0.05 5cm + 1x10−10 = 0.25 2 sec sec sec 1m sec ⎞⎞ ⎛ ⎛ x − vt ⎞ C ⎛ ⎟ + exp⎛⎜ v x ⎞⎟erfc⎜ x + vt ⎟ ⎟ Often this term is ignored C = o ⎜ erfc⎜ ⎟ ⎜ ⎜2 D t ⎟ 2 ⎜⎝ because it is typically small, but ⎝ Dx ⎠ ⎜⎝ 2 Dxt ⎟⎠ ⎟⎠ x ⎠ ⎝ we carry it⎞through here. ⎛ ⎛ ⎞ ⎜ ⎟ ⎜ 50cm − 0.05 cm 1000sec ⎟ ⎜ ⎟ including the In this case ⎜ ⎟ sec erfc ⎜ ⎟ provides a more ⎜ ⎟ 2 second term cm mg ⎜ ⎟ ⎜ ⎟ 2 0 . 25 1000 sec ⎜ ⎟ 1000 accurate result ⎜ ⎟ sec ⎝ ⎠ l C= ⎜ ⎟ 2 ⎞⎟ C is greater than ⎜ ⎛ ⎞ ⎛⎜ cm cm ⎟ ⎜ 0.05 50cm ⎟ 50cm + 0.05 1000sec ⎟ ⎜ 500 at the ⎟ sec sec ⎟erfc⎜ ⎜ + exp⎜ ⎟ 2 ⎜ ⎟ 2 average travel ⎜ cm ⎟ cm ⎜ ⎟ ⎟ ⎜⎜ 2 0.25 ⎜ 0.25 1000sec ⎟⎟ ⎟ time distance ⎜ sec ⎠ ⎝ ⎝ sec ⎠⎠ ⎝ because ⎞ dispersion ⎛ ⎛ cm2 ⎞ ⎟ ⎜ ⎟ ⎜ 2.5 mg ⎜ ⎛ 50cm − 50cm ⎞ sec ⎟erfc⎛⎜ 50cm + 50cm ⎞⎟ ⎟ dominates C = 500 erfc⎜ ⎟ + exp⎜ ⎜ cm2 ⎟ ⎝ 31.62cm ⎠ ⎟ velocity near the l ⎜ ⎝ 31.62cm ⎠ ⎟ ⎜ 0.25 ⎟ ⎜ sec ⎠ ⎝ ⎠ source ⎝ mg (erfc(0.0) + exp(10)erfc(3.1623)) = 500 mg (1.0 + 22026.46(7.743x10−7 ) = 585.3 mg C = 500 l l l 1 http://inside.mines.edu/~epoeter/_GW/22ContamTrans/C1d.xls 1000 900 800 700 600 alpha = 0.5 cm alpha = 5 cm alpha = 50 cm 500 400 300 Dispersion dominates advection when 200 V X i.e. ie 100 D 0 0 1000 2000 VX < 10 Vα + D* 3000 CONCLUSION DOES NOT CHANGE The second term is important for calculating C @ early times near the source. source plume front Normally distributed spreading degree of spreading early time later later Early time is not normally distributed because the mechanical dispersion does not move contaminants backwards so the contaminants must travel some distance from the source before the normal distribution is realized. NOTE: In fact in heterogeneous environments (which means everywhere although more pronounced some places than others) contaminants often lag behind in low K materials biasing the spreading and causing long late tails 2 Suppose a source continuously enters that uniform flow field With an initial concentration of Co=1,000mg/l pause to consider relationship of mass and concentration Mass = Conc * Volume Mass/Time = Conc * Velocity * Area = Conc * Q (Q is discharge) Mass = Conc * Q * Time Envision the source is submerged and emanates from a 0.5cm high x 1cm wide zone pause to consider the character of the source geometry v = 0.05 cm/sec dispersivity d spe s ty αx = 5 c cm dispersivity αy = 1/5 αx dispersivity αz = 1/10 αx What will the concentration be at 50 cm directly down gradient after 1000sec? pause to consider the coordinate system v = 0.05 cm sec cm m2 10000cm2 cm2 = 5cm + 1x10−10 0 . 25 sec sec 1m2 sec 2 2 cm m 10000cm cm2 1 1 = Dy = v α x + D* = 0.05 5cm + 1x10−10 0 . 05 5 sec 5 sec 1m2 sec 2 2 cm m 10000cm cm2 1 1 = 0.025 Dz = v α x + D* = 0.05 5cm + 1x10−10 2 10 sec 1m sec 10 sec Dx = v α x + D* = 0.05 ⎛ x − vx t C ⎛ C ( x, y, z, t ) = o ⎜ erfc⎜ ⎜2 D t 8 ⎜⎝ x ⎝ ⎛ ⎛ Y ⎜ ⎞ ⎞⎜ ⎜⎜ y + 2 ⎟ ⎟ erf ⎟ ⎟⎜ ⎜ ⎠ ⎠⎜ ⎜ 2 D x y ⎜ v ⎝ ⎝ 2 Dxt = 2 0.25 2 Dy 2 Dz ⎛ ⎞ ⎜ y−Y ⎟ ⎟ − erf ⎜ 2 ⎜ ⎟ x ⎜ 2 Dy ⎟ v ⎝ ⎠ ⎞ ⎞⎛ ⎛ ⎟ ⎟⎜ ⎜ z + Z ⎟ ⎟⎜ erf ⎜ 2 ⎟ ⎟⎜ ⎜ x ⎟ ⎟⎟⎜⎜ ⎜ 2 Dz v ⎠ ⎠⎝ ⎝ ⎛ ⎞ ⎜ z−Z ⎟ ⎟ − erf ⎜ 2 ⎜ ⎟ x ⎜ 2 Dz ⎟ v ⎝ ⎠ ⎞⎞ ⎟⎟ ⎟⎟ ⎟⎟ ⎟ ⎟⎟ ⎠⎠ cm2 1000 sec = 31.62cm sec x cm2 50cm = 2 0.05 = 14.14cm sec 0.05 cm v sec x cm2 50cm = 2 0.025 = 10cm sec 0.05 cm v sec 3 ⎛ x − vxt C ⎛ C ( x, y, z, t ) = o ⎜ erfc⎜ ⎜2 D t 8 ⎜ x ⎝ ⎝ 2 Dxt = 2 0.25 2 Dy 2 Dz cm2 1000 sec = 31.62cm sec x cm2 50cm = 2 0.05 = 14.14cm sec 0.05 cm v sec x cm2 50cm = 2 0.025 = 10cm v sec 0.05 cm sec ⎛ ⎛ Y ⎜ ⎞ ⎞⎜ ⎜⎜ y + 2 ⎟ ⎟ erf ⎜ ⎟⎟ ⎜ ⎠ ⎠⎜ ⎜ 2 D x y ⎜ v ⎝ ⎝ ⎛ ⎞ ⎜ y−Y ⎟ ⎟ − erf ⎜ 2 ⎜ ⎟ x ⎜ 2 Dy ⎟ v ⎝ ⎠ ⎞ ⎞⎛ ⎛ ⎟ ⎟⎜ ⎜ z + Z ⎟ ⎟⎜ erf ⎜ 2 ⎟ ⎟⎜ ⎜ x ⎟ ⎟⎟⎜⎜ ⎜ 2 Dz v ⎠ ⎠⎝ ⎝ ⎛ ⎞ ⎜ z−Z ⎟ ⎟ − erf ⎜ 2 ⎜ ⎟ x ⎜ 2 Dz ⎟ v ⎝ ⎠ ⎞⎞ ⎟⎟ ⎟⎟ ⎟⎟ ⎟ ⎟⎟ ⎠⎠ cm 1000 sec = 0cm sec Y 1cm 1cm = 0+ = 0.5cm y − = 0 − = −0.5cm 2 2 2 0.5cm Z 0.5cm = 0+ = 0.25cm z − = 0 − = −0.25cm 2 2 2 x − vxt = 50cm − 0.05 Y 2 Z z+ 2 y+ mg 0 ⎛ − 0.25cm ⎞ ⎞ ⎛ − 0.5cm ⎞ ⎞⎛ ⎛ 0.25cm ⎞ ⎞ ⎞⎛ ⎛ 0.5cm ⎞ l ⎛⎜ erfc⎛⎜ C= ⎟ − erf ⎜ ⎟ ⎟⎟ ⎟ − erf ⎜ ⎟ ⎟⎟⎜⎜ erf ⎜ ⎟ ⎟⎟⎜⎜ erf ⎜ ⎜ 8 31 . 62 cm 14 . 14 cm 14 . 14 cm 10 cm ⎠ ⎝ 10cm ⎠ ⎠ ⎠ ⎝ ⎠ ⎠⎝ ⎝ ⎝ ⎠ ⎠⎝ ⎝ ⎝ mg (erfc(0))(erf (0.0354) − erf (− 0.0354))(erf (0.025) − erf (− 0.025)) C = 125 l 1000 C = 125 mg (1)(0.0399 − (− 0.0399))(0.0282 − (− 0.0282)) l C = 0.56 mg l What do you make of the concentration relative to the C we obtained for the slug source? How much mass enters the system in 1000sec? M = CQT = CAVDT How would you go about developing a contour map of the plume? If you did not know the dispersivities, how could you use this equation to estimate them? How might you set up the problem if 8g/d arrived at the water table over a 1m2 area in an aquifer with the properties and conditions used for the example? What do you make of the concentration relative to the C we obtained for the slug source? The slug source concentration at 50cm and 1000sec is 2 orders of magnitude higher. This is intuitively reasonable. The slug added 1000mg at time zero. How much mass enters the system in 1000sec? ? M = CQT = CAVDT With the continuous source we have 1000mg/l entering an area of 0.5cm2 for 1000sec at a Darcy velocity of 0.01cm/sec so the mass is lower and mass that entered later has not had the chance to travel far. M = Conc Area VelocityDarcy Time Total Mass = 1000mg * 1liter * 0.5cm2 * 0.01cm * 1000sec = 5mg sec liter 1000cm3 On the other hand all the mass for the slug case has been dispersing for the full 1000sec. How would you go about developing a contour map of the plume? For a given point in time calculate C at many x,y,z values then map then and contour them If you did not know the dispersivities, how could you use this equation to estimate them? Collect concentration data by sampling water in the field then create a contour map using different values of dispersivity until a good match is obtained. There are some automated techniques for calibration that we will not go into here. How might you set up the problem if 8g/d arrived at the water table over a 1m2 area in an aquifer with the properties and conditions used for the example? Use formula for 2D downward spreading with a small Z, perhaps 0.0001 cm, and Y=100cm. The formula assumes the Q is the area of the source * DarcyV = 0.01cm2 * 0.01cm/sec So to get 8g/d Co needs to be = M / (Area VelocityDarcy Time) = Concentration = 8000mg / 0.01cm2 * 0.01cm * 86400sec sec = 926mg 1liter = 9.26mg cm3 1000cm3 liter 4 For a material with a half-life of 12 yrs, how much is left after 40 yrs? (Hint figure it as a % of initial mass) N = N o e ( − λt ) λ= 0.693 T1 0.693 0.693 λ= = = 0.05775yrs −1 T1 12yrs 2 2 N = N o e (− λt ) N = 1 * e (− 0.05775 yrs −1 *40 yrs ) = 0.099 or about 10% It is often said that material is essentially gone after 7 half-lives. How much is left then? N = N o e ( − λt ) 0.693 0.693 λ= = = 0.693units−1 T1 1unit λ= 0.693 T1 2 2 N = Noe(−λt ) N = 1* e(−0.693units −1 ) = 0.0078~ 0.008 *7units or less than1% 5 What is the Retardation Coefficient for a site with ⎞ ⎛ ρ Vwater Kd = 0.01ml R= = ⎜⎜1 + b K d ⎟⎟ Vcontaminant ⎝ φe mg ⎠ effective porosity of 0.3 particle density of 2.65 g/cc ρ b = (1 − 0.3) * 2.65 g 1000mg 1cc mg = 1855 cc g 1ml ml ⎛ ⎞ ρ = ⎜⎜ 1 + b K d ⎟⎟ φe Vcontaminant ⎝ ⎠ mg ⎛ ⎞ ⎜ 1855 ml ⎟ ml ⎟ = 62.8 ~ 63 0.01 R = ⎜1 + 0. 3 mg ⎟ ⎜ ⎜ ⎟ ⎝ ⎠ R= Vwater What is the Retardation Coefficient for a site with Ground water velocity = 0.05 cm/sec Contaminant velocity = 0.0009 cm/sec R= R= R= Vwater Vcontaminant Vwater Vcontaminant ⎞ ⎛ ρ Vwater = ⎜1 + b K d ⎟⎟ Vcontaminant ⎜⎝ φe ⎠ ⎛ ⎞ ρ = ⎜⎜ 1 + b K d ⎟⎟ φe ⎝ ⎠ cm 0.05 sec = 55.56 ~ 56 = cm 0.0009 sec 6