Status of SPEE Monograph 4 & The Importance of Transient Flow in Estimating Unconventional Reserves Denver SPEE chapter luncheon 14 October 2015

Transcription

1 Status of SPEE Monograph 4 & The Importance of Transient Flow in Estimating Unconventional Reserves Denver SPEE chapter luncheon 14 October 2015 John Seidle MHA Petroleum Consultants 1

2 Or How I became a modified Arps agnostic. 2

3 Status of SPEE Monograph 4 Estimating Developed Reserves in Unconventional Reservoirs 6 of chapters final edits in progress 4 of chapters revised drafts due 1 November Final manuscript to SPEE 1 December? Sister society review? Publication 1Q 2016? 3

4 Developed Reserves Historically Estimated with Decline Curves Decline curve assumptions 1. Constant bottomhole flowing pressure 2. Constant drainage area 3. Constant permeability and wellbore condition 4. Constant fluid properties 5. Well in boundary-dominated flow (BDF) 4

5 The Problem Unconventional wells can take years to reach BDF The Solution Rate Transient Analysis (RTA) 5

6 The Problem RTA requires rates and pressures The Solution Semi-empirical models 6

7 Current semi-empirical transient flow models 1. Arps 2. Modified Arps 3. Duong 4. Stretched Exponential Decline Model (SEDM) 7

8 These models can accurately forecast future well performance, even when the well is in transient flow. 8

9 Or can they? 9

10 DJ Niobrara horizontal well four analyses 1. Full history 4.7 years The Truth 2. 6 months production data 3. 1 year production data 4. 3 years production data

11 oil, water, bpd - gas, mcf Full history production data 1, time, months oil gas water 11

12 gor, scf/stb wor, stb/stb Full history GOR & WOR plot 5, ,500 4, ,500 3, ,500 2, ,500 1, time, months 0.00 gor wor 12

13 qo/(pi-pwf), bpd/psia Full history normalized rate vs material balance time plot,000 1, days 1150 days 1 0 1,000,000 0,000 material balance time, days data unit slope half-slope 13

14 oil, bpd Hyperbolic and exponential declines match boundary dominated flow exponential decline = 22.7 %/yr time, months oil b = 0.7 exp 14

15 oil, bpd The Truth Estimated Ultimate Recovery = 157 mstb 00 0 hyperbolic EUR = 163 mstb exponential EUR = 151 mstb time, months oil Arps exp 15

16 oil, water, bpd - gas, mcf 6 months production data 1, time, months oil gas water 16

17 gor, scf/stb wor, stb/stb 6 months GOR & WOR plot 5, ,500 4, ,500 3, ,500 2, ,500 1, time, months 0.00 gor wor 17

18 qo/(pi-pwf), bpd/psia 6 months normalized rate vs material balance time plot,000 1, ,000,000 0,000 material balance time, days data 18

19 oil rate, bpd 6 months Arps match & forecast b = 2 De = 82.3 %/yr time, days data Arps 19

20 oil rate, bpd 6 months Arps & modified Arps forecast De min = 8 %/yr time, years Arps modified Arps 20

21 qo/np, day-1 6 months Duong plot 1.0E E-01 a = 2.0 days-1 m = E E E E ,000,000 time, days data Duong 21

22 oil rate, bpd - cum oil, mstb 6 months Duong match & forecast,000 1, ,000 1,200 1,400 1,600 1,800 2,000 time, days oil cum oil rate fcast cum fcast 22

23 ln(qi/q) 6 months SEDM plot ,000,000 time, days data sedm 23

24 oil, bpd - cum oil, mstb 6 months SEDM match & forecast,000 1, time, days oil cum oil calc'd gas calc'd cum 24

25 oil production rate, bpd 6 months all models rate forecasts,000 1, ,000 2,000 3,000 4,000 5,000 6,000 time, days data Arps exp modified Arps Duong SEDM 25

26 cumulative oil, mstb 6 months all models cumulative forecasts 3,000 2,500 2,000 1,500 1, ,000 2,000 3,000 4,000 5,000 6,000 time, days data Arps exp modified Arps Duong SEDM 26

27 6 months EUR s & lifetimes model EUR, mstb life, yrs Arps 5,172 2,370 modified Arps Duong fails rate inc fails rate inc SEDM 43 1 hyperbolic exponential

28 oil, water, bpd - gas, mcf 1 yr production data 1, time, months oil gas water 28

29 gor, scf/stb wor, stb/stb 1 yr GOR & WOR plot 5, ,500 4, ,500 3, ,500 2, ,500 1, time, months 0.00 gor wor 29

30 qo/(pi-pwf), bpd/psia 1 yr normalized rate vs material balance time plot ,000,000 0,000 material balance time, days data half-slope 30

31 oil rate, bpd oil rate, bpd 1 yr Arps & modified Arps b = 1.2 De = ,000 1,500 2,000 time, days data Arps time, years Arps mod Arps 31

32 oil rate, bpd - cum oil, mstb qo/np, day-1 1 yr Duong plot & match 1.0E E E E-03, E-04 1, E time, days ,000 1,500 2,000 time, days oil cum oil rate fcast cum fcast 32

33 oil, bpd - cum oil, mstb ln(qi/q) 1 yr SEDM plot & match ,000,000 time, days data sedm,000 1, time, days oil cum oil calc'd oil calc'd cum 33

34 oil production rate, bpd 1 yr all models rate forecasts,000 1, ,000 2,000 3,000 4,000 5,000 6,000 time, days data Arps exp modified Arps Duong SEDM 34

35 cumulative oil, mstb 1 yr all models cumulative forecasts ,000 2,000 3,000 4,000 5,000 6,000 time, days data Arps exp modified Arps Duong SEDM 35

36 1 yr EUR s & lifetimes model EUR, mstb life, yrs Arps modified Arps Duong SEDM hyperbolic exponential

37 oil, water, bpd - gas, mcf 3 yrs production data 1, time, months oil gas water 37

38 gor, scf/stb wor, stb/stb 3 yrs GOR & WOR plot 5, ,500 4, ,500 3, ,500 2, ,500 1, time, months 0.00 gor wor 38

39 qo/(pi-pwf), bpd/psia 3 yrs normalized rate vs material balance time plot,000 1, ,000,000 0,000 material balance time, days data half-slope 39

40 oil rate, bpd oil rate, bpd 3 yrs Arps & modified Arps b = 0.95 De = 86.5 %/yr ,000 1,500 2,000 time, days data Arps time, years Arps mod Arps 40

41 oil rate, bpd - cum oil, mstb qo/np, day-1 3 yrs Duong plot & match 1.0E E E E-03, E-04 1, E time, days ,000 1,500 2,000 time, days oil cum oil rate fcast cum fcast 41

42 oil, bpd - cum oil, mstb ln(qi/q) 3 yrs SEDM plot & match , ,000,000 time, days data sedm 1, time, years oil cum oil calc'd oil calc'd cum 42

43 oil production rate, bpd 3 yrs all models rate forecasts,000 1, time, days data Arps exp modified Arps Duong SEDM 43

44 cumulative oil, mstb 3 yrs all models cumulative forecasts time, days data Arps exp modified Arps Duong SEDM 44

45 3 yrs EUR s & lifetimes model EUR, mstb life, yrs Arps modified Arps Duong SEDM hyperbolic exponential

46 Summary Unconventional well rate forecasts often difficult because-- o Multi-year transient flow DCA problematic o Only rates, no pressures available RTA problematic Solution - semi-empirical models & rates to predict performance Industry standard is modified Arps method 46

47 Summary - 2 Mature DJ Niobrara well performance forecasted with o o Arps, modified Arps, Duong, and SEDM methods 6 months, 1 year, and 3 years of production history None of the 4 methods considered-- o o o Was clearly superior Matched full history EUR Only SEDM gives constant or increasing EUR with time 47

48 Summary - 3 Estimated Ultimate Recoveries (mstb) by method & history interval Method 6 mons 1 yr 3 yrs Arps 5, Modified Arps Duong na SEDM EUR from current production = 157 mstb 48

49 Recommendations for Unconventional Wells in Transient Flow Continue to use modified Arps method to forecast future performance. Forecasts with modified Arps method are more accurate if constrained by offset and/or analogous wells o Initial decline parameters qi, b, Di o Terminal decline rate De min 49

50 Thank you!