CHAPTER-2

DRILLING OPTIMIZATION

Definition and philosophy of optimization

Drilling optimization is the logical process of analyzing effects and

interactions of drilling variables through mathematical modeling to achieve

maximum drilling efficiency. The process involves the post appraisal of offset

well record to determine the cost effectiveness of selected control variables.

Which include mud type, hydraulics, bit type, weight on bit and rotary speed.

The variables that offer the best potential for improving the drilling process

are identified. A final optimized drilling program is prepared and then will be

implemented in the field. Flexibility should be built into the program to allow

field application changes that may be dictated when unexpected problems are

encountered.

There is no such thing as a “true” optimum drilling program; invariably

compromises must be made because of limitations beyond our control that result

in something less than optimum. Perhaps it can be explained this way; for years

it has been known that rate of penetration could be increased by drilling with

water, by rotating the bit faster, and by increasing flow velocity through jets in

the bit. Lack of sufficient mechanical and hydraulic horsepower, however, often

prevents the proper balancing of variables to obtain maximum drilling efficiency.

Also, there has always been a limiting value over which an increase in rpm, weight

on bit, and pump rate does little or no good. New technology has raised these

limits, but they are still there. The limits set for drilling variables are all

influenced by the resulting bit life, the first major factor. Some are set by the

second major factor – stability of the well-bore.

Although water is the fastest drilling liquid, in many areas some colloidal

solids in the fluid are necessary to provide hole stability. In other areas, weight

on bit is limited by the deviation characteristics of the formation. Rotary speed

and pump pressure are generally limited by equipment capability and resulting

maintenance costs. Therefore, the lowest cost drilling will result when limits are

imposed that maximize not only drilling rate but also equipment life and well-

bore stability. In some cases, if well-bore stability and equipment life are

maximized, a decreased penetration rate will have to be accepted. In other

words, a balanced program must be developed – one in which the drilling

variables being considered are at their most effective level.

Developmental stages of optimized drilling

It is important to realize that optimized drilling would not be possible

today without the hard work of numerous researchers who have spent

considerable time studying the effects of drilling variables and how they relate

to each other.
 From 1937 to 1947 researchers worked on the composition and control of

drilling mud. Very little attention was given to penetration rate. The objective

was simply to get the hole drilled cased completed and production.

From 1947 to 1957 researchers focused on testing basic products used in

mud systems for effectiveness in producing desired viscosity and fluid loss

properties and to establish minimum specifications for mud materials. They also

tried to relate mud properties to well-bore stability and to hole cleaning

effectiveness.

By the late 1950’s mud chemistry had advanced to the stage where mud

properties could be controlled within specified limits by using a variety of

products available at the rig. However, mud researchers noted that even with

effective control mud properties hole conditions often were aggravated instead

of improved. It was surmised that certain formations primarily shale would

become unstable even though mud properties were within the specified

guidelines. The observations that perfect mud don’t necessarily mean perfect

holes. Prompted engineers to develop a better understanding of the relationship

of mud chemistry to wellbore behavior.

In 1959, through a series of 100 wells drilled it was shown that clay solids

had a significant effect on drilling rate number of bits and rig days. The great

improvement in overall drilling efficiency occurred as the percentage of clay

solids by volume was reduced from 6% to 0%.

Cost per foot equation:

Cost calculations are necessary when comparing less expensive milled tooth

bits with tungsten carbide insert bits. The basic cost per foot equation can be

presented as follows:

C1 = (B1 + R1 (T1 + t)) / F1

C1 = cost per foot using control bit (dollars/ft);

B1 = cost of control bit (dollars);

R1 = rig cost or operating rate (dollars/hr);

T1 = rotating time for control bit (hr);

t = round trip time (hr);

F1 = footage drilled by control bit (ft).

A second cost calculation method to determine the bit performance

required of a more expensive bit compared to a less expensive bit , based on

equal cost per foot and rate of penetration, involves the following equation:

T2 = ((B2 + R1 (t)) / (C1 (F1 / T1) – R1)

T2 = rotating time for proposed bit (hr);

B2 = cost of proposed bit (dollars).

 It is apparent from above equation that drilling cost is dependent on five

variables: rate of penetration, bit life, rig rental cost, bit cost, and non-drilling

time. Other related costs not included in the above equation are mud cost

overhead and supervision cost, and rental items.

Drilling rate equation:

R = W k N R D f / T P

R= penetration rate (ft/hr)

W= weight on bit (lb.)

N= rotary speed (rpm)

T= bit tool dullness (in.)

Df = drillability constant

k = exponent indicator of bit weight on drilling rate

R= exponent indicator of rotary speed on drilling rate.

P= exponent indicator for tooth dullness.

General guidelines and assumptions used in optimised drilling:

The concept of optimization is based on the following guidelines and

assumptions:

-All drilling variables are interrelated; changes in one variable affect all the

others.
-The type amount and colloidal size of clay solids are the factors on which all

other variables depend.

-For effective optimization, variable analysis should be approached in the

following order:

-Mud solids and type

-Mud flow properties for hole cleaning and stabilization

-Hydraulics (bit cleaning, hole cleaning stability)

-Bit type

-Weight – rotary speed conditions for bit selected.

-Consistent application of optimization techniques during drilling operations.

Mathematically, the drilling variables can be classified as alterable or

unalterable.

Alterable Unalterable

Mud Weather

Type Location

Solids Content Rig Conditions

Viscosity Rig Flexibility

Fluid Loss Corrosive Borehole Gases

Density Bottom-hole Temperature

Hydraulics Round-trip Time

Pump Pressure Rock Properties

Jet Velocity Characteristic Hole Problems

Circulating Rate Water Availability

Annular Velocity Formation to be Drilled

Bit Type Crew Efficiency

Weight-on-bit Depth

Rotary Speed
 The classification is not strict, as some of the unalterable ones may be

altered by a change in the alterable ones. For example, changes in bit type,

resulting in a faster penetration rate through a particular formation. Changing

the drilling fluid and bit type has altered the compressive and tensile strength

of the rock-drilled remains constant, but the rock’s drilling properties.

Of course, there is considerable interdependence among the alterable

variables. For instance, the type and amount of solids considerably influence mud

viscosity and fluid loss. The weight-rpm combination is interrelated; an increase

in one may necessitate a reduction in the other for smooth economical operation.

In considering which variables to choose for mathematical optimization,

experience and research suggest six: four alterable ones and two unalterable

ones.

Alterable Unalterable

Mud Formation to be Drilled

Hydraulics Depth

Bit Type

Weight-rpm

The basic interactive effects between these variables were determined by

factorial design experiments. Variable interactions exist when the simultaneous

increase of two or more variables does not produce an additive effect as

compared with the individual effects.

Cost of Drilling Fluids

Because of different drilling conditions, drilling fluid cost varies widely.

The most important factors, which affect the cost of drilling fluids, are as

follows:

-The type of formation to be drilled

-Hole size

-Total depth

-Logistics

Special drilling fluid systems are available either for reducing the severity

of these contaminations or eliminating them completely. The cost of the drilling

fluid is, however, increased as result. For example, the use of oil-base drilling

fluid in place of water base mud completely eliminates the problems of cement

and salt contaminations; however, it is very costly to change (“switch”) the mud

type.

The cost of drilling fluid increases with increasing size and depth of the well

because of the higher volume of fluid required. Logistics is also responsible for

high drilling fluid cost, because materials, which have to be transported long

distance, will cost more than those that are available locally will. Consequently,

there are no typical drilling fluid costs. For example, drilling fluid costs alone for

different wells in the west Texas area drilled to the same depths (15,000 ft)

can vary from $ 20,000 to over $ 300,000.

Drilling fluid costs can be controlled in two ways:

-As much information as possible must be obtained about the formations to be

penetrated so that early preventive actions can be taken to avoid contaminations

and subsequent costly treatments of drilling fluids. An ounce of prevention is

better than a pound of cure!

-Trained personnel must be used to supervise and direct any operations involving

drilling fluid treatments, such as treatment with chemical additives. For

example, chemical thinner is usually added if the viscosity of a drilling fluid

becomes higher than that desired, regardless of whether it reduces the

viscosity or not. Trained mud engineers, on the other hand, can properly

diagnose the situation and prescribe correct treatment. This will reduce

unnecessary treatment and lower the drilling fluid costs.

LaGrange’s interpolation formula

The following abbreviated form of LaGrange’s interpolation formula fits a

second order polynomial ( y = a + bx + c x2 ) through three data points and then

computes a value of y corresponding with a value of a selected x.

y = y1 (x-x2)*(x-x3)/(x1-x2)*(x1-x3) + y2 (x-x1)(x-x3)/(x2-x1)(x2-x3) + y3 (x-x1)(x-

x2)/(x3-x1)*(x3-x2)

Given the following table of x and y values find the value of y which

corresponds with the value of x of 26.

X 14 20 26 35 50

Y 1.0764 1.1134 ? 1.2160 1.3350

Chose the points 20, 1.1134; 35, 1.2160; 50, 1.3350 to substitute into

LaGrange’s interpolation formula . Note that the points 14, 20, and 35, could

have also been chosen.

y = 1,1134 * (26 - 35) * (26 - 50) / (20 - 35) * (20 - 50) + 1,2160 * (26 - 20) *

(26 - 50) / (35 - 20) * (35 - 50) + 1,3350 * (26 - 20) * (26 - 35) / (50 - 20) * (50

- 35)

Reduction gives the value of y corresponding to the value of 26 for x , y =

1,1525

Multiple regression with least squares:

Multiple regressions are used to find the parameters of an equation, which

causes the equation to best, represent the data. If a computer is available an

alternative, which is not associated at all with multiple regression and is very

powerful is the table look up method.

The method is illustrated with two independent variables, k, b and c. This is

a popular engineering equation. Compute k, a and b with the least squares method

for the equation.

b c
y = kx z
Begin by putting the equation into linear form by

taking logarithms.

Ln y = Ln k + b Ln x + c Ln z

For ease of presentation

Let x 1 = Ln yx2 = Ln x3 = Ln z a = Ln k

The transformed equation is

x1 = a+ bx2 + cx3

Next, the three least squares equation are solved

simultaneously

Σ x1 = a N + b Σ x2 + c Σ x3

Σ x1x2 = a Σ x2 + b Σ x2x2 + c Σ x3x2

Σ x1x3 = a Σ x3 + b Σ x2x3 + c Σ x3x3

Cramer’s rule give the solutions

D= det ( N x2 x3

x2 x2x2 x3 x2
 x3 x2x3 x3x3 )

a = det ( x1 x2 x3

x1x2 x2x2 x3x2

x1x3 x2x3 x3x3) / D

b = det ( N x1 x3

x2 x1x2 x3x2

x3 x1x3 x3x3) / D

c = det ( N x2 x1

x2 x2x2 x3x1

x2 x2x3 x3x1) / D

k is computed with the equation: k = e a

Example

Find k,b and c of the equation,

y=k*xb*z c

The linear form of the equation is

Ln y = Ln k + b Ln x + c Ln z

For ease of presentation let

x1 = Ln y * x2 = Ln x * x3 = Ln z * a = Ln k
Data computations

x y z x1 x2 x3 x1x2 x1x3 x2x2 x2x3

3 5 34 1.10 1.61 3.53 3.87 5.68 1.21 1.77

8 10 164 2.08 2.30 5.10 10.60 11.74 4.32 4.79

17 15 514 2.83 2.71 6.24 17.69 16.90 8.03 7.67

26 24 1252 3.26 3.18 7.13 23.24 22.67 10.62 10.35

37 35 2594 3.61 3.56 7.86 28.39 27.95 13.04 12.84

50 48 4804 3.91 3.87 8.48 33.16 32.82 15.30 15.14

n = 6.00

Σ x1 = 38.34 D = 2.06

Σ x2 = 16.79 a = 0.74

Σ x3 = 17.22 b = 0.85

Σ x1x2 = 116.95 c= 1.14

Σ x1x3 = 117.76 k = 2.10

Σ x2x2 = 52.52 ; Σ x2x3 = 52.57 ; Σ x3x3 = 52.95

The sought after equation becomes Y= 2,10 * x 0,85 * z 1,14

Example

The hydraulic equation is Pf = jqm

Find: j and m

Linear form : Ln Pf = Ln j + m Ln q

Let: x1 = Ln Pf , x2 = Ln q n = Lnj

q Pf x2 x1 x1x2 x2x2 y % error

569 1085 6,34 6,99 44,34 40,24 1084 0,12

 455 738 6,12 6,60 40,42 37,46 776 -5,18

347 505 5,85 6,22 36,41 34,21 518 -2,59

285 444 5,65 6,10 34,46 31,95 386 13,01

228 272 5,43 5,61 30,44 29,48 277 -1,79

171 172 5,14 5,15 26,47 26,44 180 -4,79

n = 6,00 ;∑ x1 = 33,67; ∑ x2 = 34,54 ;∑ x1x2 = 212,53

∑ x2x2 = 199,78

D = 5,89

a = -2,48

The values of j and m are computed to be

j = 0,084

m = 1,492 and equation is : Pf = 0,084 * q 1,492

Confidence Lines

Confidence lines area computed and drawn to provide the following type of

answer :

" If 7 " casing is selected 95 % of the pipe will have an expected maximum

diameter of 7.012 " and a minimum expected diameter of 6.889 " and an

expected diameter of 7.001".These are often called the maximum , minimum and

best values.

Draw to confidence and the least squares lines of the equation.

y=a+b*x
Begin by computing a,b and c

b = N ∑ xy - ∑ x * ∑ y / N ∑ x2 - (∑ x )2

a=∑y/ N–b∑x/N

χ=∑x/N

where:

x, y : data point pairs.

N : number of data point pairs.

The least squares line is given by the equation = y = a + b x

Ascertain the value of the student's 't' distribution

Example

If 95% confidence is desired and N = 4

then Γ = 1 – 0,95

and then tn = 2,1 - Γ / 2 * 0.975 = 4.30 (value depends on number of data points)

The two confidence lines as shown in the sketch line are:

y = a + b x +tS upper line

y = a + b x +tS lower line

Where S is given by the equation below

S = [1/N + (x -χ)2 / ∑ (x - χ)2]1/2 * [∑ (y – a – b x)2 / N-2]1/2

Example
 x data y data x * y x * x (x - c)2 (y – a – b x)2

2,106 2,321 4,888 4,435 5,609 0,046

3,406 3,159 10,760 11,601 1,141 0,045

4,692 4,264 20,007 22,015 -0,047 0,129

5,432 5,692 30,919 29,507 -0,947 0,122

6,736 6,621 44,599 45,734 5,115 0,000

N = 5,000 ∑ x = 22,372

x = 4,474 ∑ y = 22,057

y = 4,411 ∑ (x - x )2 = 12,830

∑ xy = 111,172 ∑ (y – a - bx)2 = 0,342

∑ x2 = 112,931 a = 0,059 b = 0,973 n -2 = 3,000 1 - χ / 2 = 0,975 student t =

3,180

Selected pairs for plotting chart

x0 y0 y0+ Y 0- S

2 2,004 2,888 1,121 0,278

3 2,977 3,630 2,325 0,205

4 3,950 4,451 3,449 0,157

5 4,923 5,428 4,417 0,159

6 5,895 6,558 5,232 0,208

7 6,868 7,765 5,972 0,282

The cost per foot equation is used for the comparison of alternative equipments,

chemicals, and procedures for the drilling of a formation or an interval. The

comparisons which are often called break-even calculations and are usually

between drill bit types or manufacturers.

 C=
[Bit + Tools + Mud + ((DrillTime + Trip + Lost ) × (Rig + Support + Tool Re ntal ))]
DrillingRate × DrillingTime

C = Cost per foot for the interval of concern $/ft

Bit = Cost of delivered bit at the drill site $

Tools = Cost of tools or repairs the tools $

Mud = Cost of mud to drill the interval $

Drill Time = Time required to drill the interval, bit life hr

Trip Time = Time to pull and run a bit hr

Lost Time = Time chargeable to non-drilling task hr

Rig = Contract rental rate of a rig $/hr

Tool Rate = Rental of tools $/hr

Drilling Rate = Average drilling rate over the interval ft/hr

EXAMPLES

a) Calculate cost per foot by using the following data

Bit = $ 1500 Tool = $ 300 Mud = $ 2500

Drill Time = 40 hr Trip = 8 hr Lost = 0.5 hr

Rig = 600 $/hr Support = 200 $/hr Tool Rental = 10 $/hr

Drilling Rate = 15 ft/hr

C=
[1500 + 300 + 2500 + ((40 + 8 + 0.5) × (600 + 200 + 10))] = 72.6$ / ft
15 × 40
 b) Calculate cost per foot if company use a bit costs $ 1000 more and

mud $ 2000 more which result to drill 15 % faster and 15 % longer

Bit = $ 1500 + $ 1000 = $ 2500

Mud = $ 2500+ $ 2000 = $ 4500

Drill Time = 40 hr × 1.15 = 46 hr

Drilling Rate = 15 ft/hr × 1.15 = 17.25 ft/hr

C=
[2500 + 300 + 4500 + ((46 + 8 + 0.5) × (600 + 200 + 10))] = 64.8$ / ft
17.25 × 46

TIME VALUE OF MONEY (ref1)

Because interest is paid on savings, the value of money varies with time.

So the time value of money can be found by using the following formula.

n× q
  r 
F = P × 1 +  
  q 

F = Future value of the money $

P = Present value of the money $

r = Rate of return fraction

n = Number of years which have transpired years

q = Number of payments per year

EXAMPLE

Find the value of $ 2000 after 10 years period if the rate of return or

interest rate is 25 % and each year has 3 payments.

n×q 10×3
  r    0.25 
F = P × 1 +   = 2000 × 1 +   = 22074$
  q    3 

EXAMPLE

Find the arte of return or interest rate if the future worth of the

present $ 1000 is $ 9000 after 15 years later and a) each year has one payment

and b) each year has 2 payments?

n× q 15×1
  r    r 
F = P × 1 +   ⇒ 9000 = 1000 × 1 +   ⇒ r = 16%
  q    1 

n× q 15× 2
  r    r 
F = P × 1 +   ⇒ 9000 = 1000 × 1 +   ⇒ r = 15%
  q    2 

EXPECTED VALUE METHOD

Expected value method provides a means of reaching a decision based on

expected costs and the probability of their occurrence. The fundamental form

of the expected value equation is given below.

 EV = (C1 × P1 ) + (C 2 × P2 )

EV = Expected value $

C1 = Cost of the first event $

P1 = Probability of the first event fraction

C2 = Cost of the second event $

P2 = Probability of the second event fraction

EXAMPLE

Calculate the expected value of 4 inch and 5 inch DP by using given below

as the hole aimed to be drilled to 12000 ft.

DP Size inch Delivery Rental Drill Rate Probability %

$ $/day ft/hr

4 1100 850 300 15

330 85

5 1300 1000 320 35

350 65

EV = (C1 × P1 ) + (C 2 × P2 )

EV4 = (C 41 × P41 ) + (C 42 × P42 )

EV4 = ((12000 300 )(850 24) × 0.15) + ((12000 330 )(850 24 ) × 0.85)

EV4 = (1415 × 0.15) + (1288 × 0.85) = 1307$

EV5 = (C 51 × P51 ) + (C 52 × P52 )

EV5 = ((12000 320 )(1000 24 ) × 0.35) + ((12000 350 )(1000 24 ) × 0.65)

EV5 = (1563 × 0.35) + (1429 × 0.65) = 1476$

 So expected saving if 4” drill pipe is selected is 169$

LAGRANGIAN MULTIPLIER (ref1)

If there exists an objective and a constraning equation the Lagrangian

multiplier method can be applied in order to find the maximum and the minimum.

The following formula implies the Lagrange equation.

 ∂f ∂g ∂f ∂g 
 ∂L ∂w − ∂w ∂L  = 0
 

f = objective equation g = constraint equation

w = parameter to be determined

L = parameter to be determined

EXAMPLE

The perimeter of a rectangular mud pit is 160 ft with a depth of 10 ft. In

order to maximize the volume of the mud pit what should be the length, L and

width, w?

Objective equation f = 10 × w × L

Constraint equation g = 2 w + 2 L − 160

 ∂f ∂g ∂f ∂g 
Lagrange equation  ∂L ∂w − ∂w ∂L  = 0
 

∂f ∂f
Partial derivatives = 10w = 10 L
∂L ∂w

∂g ∂g
=2 =2
∂L ∂w

Solving Lagrange (10w × 2) − (10 L × 2) = 0 ⇒ w = L

Length 2 L + 2 L = 160 ⇒ L = 40 ft

Width 2 w + 2 w = 160 ⇒ w = 40 ft

Maximum Volume V = 10 × w × L

V = 10 × 40 × 40

V = 16000 ft 3