WO2017146592A1 - Hoisting system and method for operating the same - Google Patents

Abstract

There is described a hosting system (1 ) for a drilling rig, the hoisting system comprising: - a wire rope (2); - a winch (4) for pulling in and letting out said wire rope (2); - a support structure, such as a derrick; - a wire rope guiding sheave (6) connected to the support structure, said wire rope guiding sheave (6) being provided between said winch (4) and a load suspension member along the wire rope (2); and - a first stabilizer (8) for damping lateral vibrations of the wire rope (2) between said winch (4) and said wire rope guiding sheave (6), wherein said first stabilizer (8) is provided closer to the wire rope guiding sheave (6) than to the winch (4) along the wire rope (2), There is also described a drilling rig comprising such a hoisting system (1 ) as well as a method for operating the hoisting system (1 ).

Description

HOISTING SYSTEM AND METHOD FOR OPERATING THE SAME

Background

Drawworks is a common name for the actuator being used for moving the drill string up and down in the well. It consists of a multi-layer drum powered by electrical motors, a drill line running from the drum over a fast sheave near the top of the derrick, then strung several times between the crown block and the travelling block and then back down to a dead line anchor locking it to the rig structure. The drill line is a steel wire rope having a nominal diameter of typically 1.5 to 2 inches. The fast line is the part of the drill line that runs between the drum and the fast sheave. It has a fixed length, typically 50 m, and it is more and less free to move transversally. Both the lower drum end and the upper fast sheave end represents dynamically fixed ends, making the fast line behave very similar to a string of a musical string instrument. The dimensions, resonance frequencies and excitation mechanisms are quite different, though. While a musical instrument string is set into vibrations by plucking it (guitar), by a hammer (piano) or by a bow (violin), the fast line is excited by lateral motions at the drum end and also by a non-linear coupling between dynamic tension force and transversal vibrations. Because there is very little damping of transversal drill line vibrations, large amplitudes can be the result if one of the drum spooling harmonics coincides with the natural line frequencies. In extreme cases the fast line can hit surrounding equipment and cause damages, also to itself. Large transversal vibrations are highly undesirable also because they cause excessive wear of the line and of the fast sheave. In extreme cases erroneous drum spooling can occur during hoisting operation. If the line jumps over one wrap during hoisting, it can lead to severe line damage when the next layer is filled.

To avoid large and damaging transversal line vibrations it has become common practice to install one or, and in some cases, two so-called fast line stabilizers. Such a devise is a movable mass with guide rollers mounted some distance above the drum. It is relatively free to move laterally by means of wires running over guide sheaves. Its mass and the friction in the guiding system are indented to reduce resonant vibrations and thereby hindering damages to the line or to other equipment. However, experience shows that the current type and placement of stabilizers is not always very efficient and that severe damages can still occur, especially when the tension is low and the hoisting speed is high. The main disadvantages of current state of the art are listed below. The radial and axial directions refer to the drum axis. 1. The stabilizer guide system represents virtually no damping in the radial direction of transversal vibrations.

2. The damping in the axial direction is dominated by non-linear Coulomb friction in the

sheaves of its guide system. The ratio of friction force to inertia forces for this kind damping decreases with the speed amplitude. This feature is far from optimal because the largest vibrations in general need the highest damping.

3. Poor damping of axial vibrations also means higher fluctuations of the fleet angle, defined as the angle between the fast line tangent and the perfectly straight line between the drum center and the fast sheave. Higher fleet angle fluctuations imply excessive forces and wear on the sides of the sheave groove.

4. The stabilizer system occupies a lot of space in the derrick because the comprehensive guide system with tensioning wires and sheaves need space to accommodate the large axial motion.

5. The system is relatively expensive to install and to maintenance.

6. Their poor performance and the risk of erroneous drum spooling sometimes cause the maximum hoisting speeds to be limited, which in turn, means poor rig performance and high costs.

The invention has for its object to remedy or to reduce at least one of the drawbacks of the prior art, or at least provide a useful alternative to prior art. The object is achieved through features, which are specified in the description below and in the claims that follow.

The invention is defined by the independent patent claims. The dependent claims define advantageous embodiments of the invention.

In a first aspect the invention relates to a hosting system for a drilling rig, the hoisting system comprising:

- a wire rope;

- a winch for pulling in and letting out said wire rope,

- a support structure, such as a derrick;

- a wire rope guiding sheave connected to the support structure, said wire rope guiding sheave being provided between said winch and a load suspension member along the wire rope; and

- a first stabilizer for damping lateral vibrations of the wire rope between said winch and said wire rope guiding sheave, wherein said first stabilizer is provided closer to the wire rope guiding sheave than to the winch drum along the wire rope. In a second aspect the invention relates to a method for damping lateral vibrations in a wire rope by means of a hoisting system according to the first aspect of the invention, the method comprising the steps of:

- measuring the rotational speed of a winch drum of the winch;

- measuring the tension of the wire rope between said winch and said wire rope guiding sheave; and - calculating optimized motion characteristics, for reducing said lateral vibrations, of one or more dampers of said first stabilizer based on said measured rotational speed and tension.

Herein the wire rope will also interchangeably be referred as a line or drilling line, while the portion of the wire rope between the winch and the wire rope guiding sheave will also be referred to as the fast line. The wire rope guiding sheave will also be referred to as a fast sheave. All this is in accordance with common terminology as used in a traditional draw-works on a drilling rig. However, the invention should not be construed as limited to traditional draw-works, but could also be used in other hoisting systems using single layer winches and/or hoisting systems without the traditional stringing between the crown block and travelling block but using only a mechanical advantage of 2- 3 or even direct drive between the winch and the load. One such hoisting system is disclosed in WO 2014/209131. Still, some of the advantages of the invention discussed herein may be more pronounced when used in a traditional draw-works as described above, in particular because the lateral fast line vibrations are more pronounced when the speed of the fast line is high.

In the following is described the theoretical basis for the invention and an example of a preferred embodiment illustrated in the accompanying drawings, wherein: shows simulated normalized drum-induced deflections, and their derivatives, as a function of winch drum rotation angle;

Fig. 2 shows the simulated normalized combined winch drum deflections as a function of winch drum rotation angle and the winch drum harmonics at the third layer; Fig. 3 shows the simulated modulus of the normalized wire rope impedance as a function of frequency without and with different stabilizer provided near the winch drum;

Fig. 4 shows the results of similar simulations as in Fig. 3 but with the stabilizer provided near the wire rope guiding sheave/fast sheave;

Fig. 5 shows the results of similar simulations as in Fig. 4 but with different stabilizer damping characteristics;

Fig. 6 shows simulated modulus of the normalized wire rope impedance as a function of frequency;

Fig. 7 shows schematically a hoisting system according to the first aspect of the present invention as well as a more detailed view of one embodiment of a stabilizer as used in the hoisting system;

Fig. 8 shows schematically one embodiment of damper for a stabilizer as used in a hoisting system according to the present invention; Fig. 9 shows simulation results of the axial wire rope deflection and fleet angle, respectively, at the winch drum and near the wire rope guiding sheave as a function of time without any stabilizer; and shows similar simulations as in Fig. 9 but with a stabilizer placed near the wire rope guiding sheave;

.Theoretical basis

The theory description below is included to provide credibility and to justify that the new stabilizer concept together with its new location does represent a significant improvement as compared with current state of art. The theory first describes the main excitation mechanism for transversal vibrations. Next it describes how the line dynamics is affected by a stabilizer, both its mass, damping coefficient and location. The line dynamics are studied by two models, one simplified linear model suitable for frequency analysis and one full simulation model suitable for studying dynamics in the time domain. As a typical but non-limiting example we shall study a standard multi-layer drum having so-called Lebus grooves. These grooves force the line of the first layer to follow a path of discrete but smooth steps, not a helical path often used for single layer drums. Two times per revolution (once every 180 degrees) the line is shifted half a pitch during a transition length of typically 30 drum degrees or 10 line diameters. Between these transition sectors the grooves are cylindrical thus representing no dynamic deflections. At the end of the first layer the radius is increased during another transition angle up to the second layer radius. When the line is spooled on the 2^nd layer, the axial motion is a kind of mirror of the 1^st layer motion. Because the line has to cross over the underlying line, the fast line is also forced to make a temporary, bell shaped radius increase at every cross-over. A detailed description of the drum-induced line motion can be given by assuming that

1. the line is perfectly circular with a radius denoted by r ,

2. the pitch, which is the c-c distance between neighbor wraps, is slightly larger than the diameter: p > d - 2r ,

3. the deflection change rate through the transition sectors is continuous and smooth. We shall need a help variable representing the angular position of the contact line relative to the center of the underlying rope. This angle varies from γ = -γ₀ to γ - χ₀≡ sin^~' (0.5/?/ d) during the cross-over interval. One possible but no-limiting choice for the variation of this contact position is a sinusoidal contact angle variation such that

where θ₀ denotes the drum rotation angle of the center of crossover and θ_χ is the angular crossover length. The resulting cross-over deflections in respective axial and radial directions are

X = X_c + r sin γ (2)

R - j?_j + rcosy (3)

Here X_c is the position at the center of the cross-over while R_{ refers to the center radius of the first layer.

The formulas above apply inside the cross-over intervals

< 0.5Θ_Χ only, but they can easily be generalized to any drum angle of any layer. The axial iine displacement at the drum is basically equal for each layer except that even layer numbers represent mirror motion of the odd layers, in contrast, the radius bump height increases with the layer number. Explicitly, the cross-over radius of any layer number, i, , can be expressed by

R ~ Ρ. + (/; - l)r cos γ (4)

The radius changes at layer shifts can be modelled similarly by assuming that the radius in the transition from layer i, - 1 to layer i, follows the function

R ~ R, + r cot γ_ϋ sin γ + i,r cos γ_ϋ (5) in this equation the drum angle defining the help variable γ is the layer shift angle $ ₍. ,

It is convenient to define normalized motions and their first and second order derivatives, primes denoting differentiation with respect to Θ :

X' Απ'

-smy (10)

The last expressions for the derivatives are valid only inside the cross-over intervals. Outside these intervals the normalized motions are constant while the derivatives are zero, with the exception of the layer shifts. Here the first and second derivatives are obtained by differentiating equation (5).

The above functions are plotted versus the drum angle in three subplots in figure 1. The x-axis represents the angle in radians beyond the filled 2^nd layer angle, here denoted by #₂₃ . The three subplots show the normalized deflections, their first derivatives and their second derivatives, re- spectively. The first derivatives represent the angular deviation from the tangent line while the second derivatives represent the line curvatures, or more precisely, the curvature deviation from the base curvature. The axial deflection steps are slightly larger than unity (line radius), reflecting the fact that the pitch in this example is chosen to be 2.5% larger than the line diameter. Also we have assumed that both the layer shift interval and the crossover intervals are 36 degrees. The physical transversal speeds and accelerations of the line at its lower end can be calculated from the time derivatives of the deflections: v_r = R = rQR_n ^* (12) ν_Λ = Χ = τΟΧ_η' (13) e_r = j? = ΓΩ²Λ_β* (14) a_x = X = τΩ²Χ_η" (15) where

Ω≡θ = ^- (16) R

is angular drum speed and v_l is the line speed. This line speed simply equals the block speed times the number of line parts. As a numerical example, consider a drum of radius of R = 0.75/« and a 1 ¾ " line ( r = 0.0222 m ) running at a typical maximum line speed of v. = 25 m/s . The corresponding angular speed is Ω = 333rad I s , corresponding to a fundamental cross-over frequency

is f_xo = 2Ω/2Τ = 10.6Hz . Maximum radial peak speed occur at the layer shift where R,' _max - 4.5 and equals v, _max = 3 A m/s . Maximum radial acceleration at the 3^rd layer crossovers

( R'_max = 26.3 ) is a_{r max} - 650 m/s² « 66g . These peak acceleration levels can be characterized as extremely high and could possibly be a cause of the high failure rates and maintenance costs of the current fast line stabilizers. Notice also that the peak acceleration increases proportionally to the line speed squared and inversely proportionally to the crossover length squared. In mathemati- cal terms, _{r !nm}∞ r vf /(R0_X)² .

Anticipating the suggestion below of absorbing the line motion by dampers mounted in directions of characterized by angles ± π 14 relative to the pure radial direction, is useful to study the combined motion R_n ' cos(;r/4) + X_n ' sin(/r/4) = (R_n'+X_n ' )/ 2 and its frequency components. Reference is made to figure 2. The upper subplot shows this combined speed versus the drum angle over the same drum angle interval as in figure 1. The lower subplot shows the corresponding harmonics spectrum of the third layer ( Θ > θ₂₃ + θ_χ ). Notice that the combined speed has a lower peak (3.2) at the layer shift than the pure radial speed and that the individual harmonics components are much smaller in amplitude (in the order 0.5) than the peak speed itself (approximately 3.4). The amplitudes are relatively constant up to the 10^th harmonics, then they fall off and are quite small for the 20^th harmonics and above. The harmonics spectra for the normalized deflection and acceleration are not included here but they can easily be obtained by respective dividing and multiplying the speed spectrum by the harmonics number. We then find that the harmonics components of the deflection fail rapidly off with frequency while the acceleration can have large high frequency components. The maximum acceleration components are found around the 10^th harmonics for this par- ticular case where the crossover angle covers 1/10 of a revolution.

The spectra discussed above apply if the drum is rotated at a constant speed on the 3^rd layer. We see from the upper subplot in figure 2 that the layer shift is a non-periodic pulse representing a continuous spectrum with all kind of frequencies.

It should also be mentioned that there are other type of excitation mechanisms that can put the fastline into transversal vibrations but are not included in the mathematical models here. One such mechanism is the non-linear coupling between dynamic tension force and transversal vibrations. When the line deflects laterally, it is slightly longer than the perfectly straight line and this elongation causes a variation in tension. The coupling also works the other way around, so that a change in tension affects the deflection. This mechanism can be quite strong, especially when the fast line is statically inclined (deviates from perfectly vertical) and has a low tension force. The gravitation induced, static deflection is nearly proportional to the inclination angle but inversely proportional to the tension. When the tension increases rapidly from a low value, for instance when stopping a downwards motion of an empty travelling block, then this mechanism can give rise to large transi- ents line vibrations. Because the shape of the static deflection is very similar to the amplitude function of the fist harmonic vibration, it is mainly the first harmonic vibration that is excited by such stops of an inclined fast line. Other excitation mechanisms for transversal line vibrations are external derrick motion (applicable for floating rigs) and wind induced line forces. The vibrations induced by ail these alternative excitation mechanisms will be equally well dampened by the new stabilizer as the drum generated vibrations discussed below.

The next step is to study how the fast line responds to drum-induced motions, and in particular, how the placement and motion characteristics of a stabilizer influence the response. The study consists of two different approaches, the first being a linear, frequency-based study and the other a time based simulation model. The second approach is a numeric and non-linear simulation model being discussed at the end of this section.

The first approach is based on the following simplifying assumptions.

1. The fast line tension force, T , is constant, independent of time and position

2. The damping is low but linear and comes mainly from internal, bending induced friction.

3. The fast sheave is either laterally fixed or it has linear response characteristics.

4. The longitudinal fast line speed, v_; , is small compared with the transversal wave propagation speed c =

, m being the specific mass (mass per unit length).

Their validity is briefly discussed below. The fast line tension normally varies much during different operations, especially during accelerations of the drum motion. However, in a wide range of operations the tension is fairly constant, at least on time scales comparable with the lowest natural. The line weight will make the tension increase with height, but normally the tension force is much higher than its weight (gravitation force). Therefore first assumption is fairly well satisfied, at least for a wide range of operations. The second assumption is more questionable, mainly because the internal bending resistance is not linear but a Coulomb type friction. However, as long as it is relatively small the linear approximation is far better than neglecting damping. The third assumption is also well satisfied in normal cases. Finally, the validity of the last assumption is justified through the following numerical example. An empty block with a top drive have a weight of typically 600 kN, corresponding to a line tension force of T = 50 kN if the blocks are strung with 12 line parts. A standard 1 ¾" line with a specific mass of m = $A4 kg/m will then represent a transversal wave propagation speed oi c = 11 mj . In comparison the typical maximum line speed

is v_; = 25 mjs = 0.32c . It can be shown mathematically that the effect of the longitudinal line speed has a 2^nd order effect on the natural frequencies, which are all lowered by the factor 1 - vf /c « 0.9. We can therefore conclude that the last assumption is fairly well satisfied even with the combination of a high longitudinal line speed and a low tension force

The classical equation of motion for a loss-less stretch string is d²u _^ d²u

m— r- = T—— (17 dt² dz²

where u is the transversal deflection (in any of the two possible directions), t is the time variable and z is the longitudinal position along the straight line between the end points. The same equation applies also for the transversal speed, which is the time derivative of the deflection:

v = d jdt . For convenience, and for avoiding singularities in the response functions, we shall need some kind of damping along the line. For this size of wire ropes the damping is probably dom- inated by bending resistance, which is a kind of internal friction being proportional to the rate of curvature change. We shall therefore study the following equation of motion.

Here r is an internal friction parameter having the dimension of time. The factor 2 is included for convenience, to achieve some simplification in subsequent equations. This kind of damping term gives a damping rate with drops rapidly with frequency. It is possible even within the frame of linearity to change the line damping characteristics by considering r as a function of frequency and tension. But here we shall assume it is constant.

Because of the assumed linearity, Fourier analysis applies. It means that any solution can be described as a sum of independent frequency components. The general mono-frequency solution to this equation is the sum of two damped waves represented by v = ae^j'"^~jh + be^l6*^+jk: (19) Here a and b are arbitrary complex amplitudes of the waves travelling in positive and negative z- direction, respectively. Furthermore, j— V-I is the imaginary unit, ω is the angular frequency and k is the wave number, here assumed to have a positive real part. The two last parameters are related to each other through the equation c²co² - k²(l + ]2τω) ^r_j)

This equation is obtained by putting the first term of the general solution into the equation of motion. If the damping is small, 2τω « 1 , then the wave number can be approximated by

where k₀ = cojc is the !oss-iess wave number. In the following we shall, for convenience, omit the common time factor exp(yVrf) .

Energy transmission along a transmission line can always be expressed as the product of a forcing variable and a motion variable. Sn this case the motion variable is the transversal speed while the forcing variable is the line shear force expressed by

Sn the last expression we have introduced the characteristic impedance l

ζ _— = (1 - jT_bco)s!Tm = (1 - jx_bco)mc (23) ω

For a lossless line the characteristic impedance is purely real it is also convenient to define general mechanical impedance as the ratio of transversal force to transversal speed, applicable both to a specific location of the line and to a lumped element connected to the line. A wide class of lumped impedance elements can be characterized by its mass , its damping coefficient B and the stiffness S of an optional spring. The complex Fourier representation of such an impedance element is

Z_s = jcoM + Z (24) where

S

Z„= B + (25)

If an elastomer (rubber) type coupling is used, it can, to a fairly good approximation, be modelled with stiffness-proportional damping, that is, 5 = 0 and S = S₀ ^■ (1 + _,/^'??) . Here η is a hysteresis loss factor, in the range 0.15 - 0.3 for typical elastomer qualities.

The above single degree of freedom (DOF) lumped impedance can be generalized to include multiple masses and coupling elements between the line and a fixed point. It can be shown that the impedance of an n-DOF system can be found by the following recursion formula. Z, _t = icoM, + (Z„ Z.._.^"' ) ^" (26) Here Z_{c i} is the coupling impedance between masses M, and and the higher index is closer to the line.

The coordinate system is here chosen so that the longitudinal location variable z equals 0 at the fast sheave and - I at the drum. In the general wave solution (10) above, the amplitudes a and b therefore represent the amplitudes of the incident and reflected waves, respectively. It can be shown that the reflection coefficient at the top line end ( z = 0 ) can be expressed by the ratio

£ ^ ^-j£ (27) where Z₀ represents the impedance of the fast sheave. Normally, when the fast sheave support is very stiff, this impedance is close to infinity, making b = -a . If there are no impedance elements between the line position z and the top end, the impedance at the location z equals

Z(z) = ₌ ^' ^- ^ ₌ ¾ oos(fej - f sk(fe _(2g v(z) m~^ ^h ^{f i} ¾sin(&$

These expressions are valid even when linear damping makes the wave number complex. A useful special case is when the fast sheave is rigidly supported ( |Z₀| =∞ ) and there are no stabilizers. The line impedance at the drum end then equals Z_d Z(-L = ^^ = -jcat kL) (29) j sm(kL)

For low damping values the modulus of the normalized drum impedance,

, has high peaks at the discrete frequencies ^ irr (30) i being a positive integer. We identify these frequencies as the natural (harmonic resonance) fre- quencies of the classical lossless string. With a small damping this normalized impedance can reach very high values, maybe hundred or more. The implication is that the dynamic transversal deflection and speeds can reach very high values if one the excitation frequencies, which are even multiples of the drum rotation frequency, matches one of the above string resonance frequencies.

Another interesting special case, although being difficult to implement physically, is when the top impedance matches the characteristic impedance: Z„ = ζ . Then the reflection coefficient equals zero and the impedance at the drum also equals: Z_d = ζ . This is called perfect impedance match and is consistent with the foiiowing ruie in physics, If the end impedance of a transmission line equals its characteristic impedance, then the transmission iine behaves as a reflection free, semi- infinitely long line.

The results above can easily be generalized to cases where one or more lumped elements are attached to the iine. Assume that one stabilizer with a lumped impedance Z_s is placed at a distance / from the top. The impedance just below the stabilizer equals the sum of this lumped impedance and the line impedance just above the stabilizer, in mathematical terms:

This impedance can now be regarded as the end impedance seen by the line below it. if more than one stabiliser is used, the above formula must be used recursively by the following formula.

where /₍. is the distance between stabilizers - 1 and i , and Z..(i) is the lumped impedance of stabilizer number (not be confused with the partial impedance in equation (26)) . if there is only one stabilizer, the line impedance at the drum can be expressed by

if the fast sheave impedance is infinite, the above expression equals

This expression is valid also for a muiti DOF damper where Z_s - Z_{s n} and Z_{s n} is found from the recursive formula (26) above. Before discussing special cases it is worth noticing that the normalized impedance Ζ_ά /ζ represents a response function describing how the dynamic line vibration is ampiified in comparison with a line with no reflection at the top. More specifically, the dynamic fleet angle amplitude (in radians) at the drum resulting from an axial excitation speed v_g equals

, F(-L) Z . v

c ζ c

It is not easy to see directly from the formula how this impedance varies but numerical examples below show that a stabilizer will change the drum impedance spectrum rather dramatically. Key parameters for most of the subsequent examples are the following. The special tension force is chosen for convenience, so that the wave propagation speed and the harmonic frequencies become round numbers. This tension force represents a typical line load when the hook load includes a drill string of moderate weight.

T - MAkN tension force

L = 50m fast line length

2r = 0.0445 m line diameter (1 ¾")

m = 8.44 kgm^~[ specific line mass

T = 2 · ] 0 ⁴ s internal bending loss time constant

c

00ms ' wave propagation speed

ζ = m - 844A¾m^{" 1} characteristic impedance (lossless case)

Data for an optional (1-DOF) stabilizer

M = 80kg mass {12% of the fast line mass ml = 422kg )

B = 422 N$m^~ - 0.5 ζ damping coefficient (optional)

/ = 45 w location ( OA L above drum)

The modulus of the normalized line impedance at the drum position is visualized in figure 3 for three cases. The first case, represented by the dotted curve, is without any stabilizer. The resonance peaks in this case are integer multiples of the fundamental resonance frequency

/, = Ω,/2π = 1.0Hz . The reduction of the peak heights with increasing frequency is a consequence of the bending induced damping, which increases rapidly with frequency. The logarithmic amplitude scale is used for visualizing the extreme differences in the dynamic line response. With the assumed internal line damping the response spans 5-6 orders of magnitude, from the maxima at the resonance frequencies to the minima at the anti-resonance frequencies. At the latter fre- quencies the force response is nearly zero, meaning that the line behaves as if it is not present.

The second case, the dashed-dotted curve, represents a stabilizer without damping. The stabilizer mass changes the line response dramatically. The most pronounced difference is that a new wide resonance peak appears at a frequency slightly higher than the 10^th harmonics of the stabilizer-less spectrum. This peak is the first one in a series of new, regularly spaced frequency peaks being the result of the mass being so close to the drum and so far away from the fixed fast sheave. We also see that the lowest line resonances are still there but they are shifted slightly to lower frequencies.

The third case, represented by the solid curve, is with added stabilizer damping. The extra damping reduces the lowest resonance peaks substantially, almost by a factor 50. In contrast the damping has only a small effect on the response above the new resonance peak. This can be explained by the fact that the high line damping makes the wave being reflected from the fast sheave is heavily attenuated when it reaches the stabilizer.

Case 2 (without damping) fairly well represents the impact on radial vibrations by the kind of stabilizer being used today. This is because there is virtually no friction restricting the stabilizer motion in the radial direction. The last case with damping is intended to simulate the action of current stabi- lizers on axial vibrations, although the effective level of the actual Coulomb friction is very uncertain. It is interesting to see the effect of flipping the stabilizer position around the line centre. In the reference example above the distance from fast sheave to stabilizer equals / = 0.9L . Figure 4 shows the corresponding impedance spectra when / - 0.1Z , which is much closer to the fast sheave. In this position the stabilizer has a much different effect on the line dynamics. The pronounced im- pedance peak between the original 10^th and 11^th harmonic observed with the low stabilizer is not seen any more. Without damping the new impedance spectrum is very much like the original spectrum, except that the frequencies are shifted slightly. When damping is added, the peak heights of the lowest resonances are very much reduced, to approximately 1 % of the original height. Notice that the 10^th harmonic peak, which has a node at the stabilizer position, is virtually not affected by the stabilizer at all.

It can be seen that for frequencies making

« 1 the line impedance above the stabilizer is Z(-/) = -]ζ cot(&/) « T l( j(d) . The last approximation equals the impedance of a linear spring with stiffness S - Tjl . We can therefore deduce that perfect impedance match is obtained at the angular resonance frequency of the stabilizer ω = ^JT/(IM) if the damping matches the charac- teristic line impedance, that is, B - ζ = sjT/m . Because the line has many resonance frequencies it may be more optimal to increase the damping beyond the characteristic impedance. The sacrifice of total impedance match at one particular frequency is outweighed by the better damping performance over a wider frequency band.

Reducing the stabilizer inertia is another way of reducing the resonance in the band between probably be de¬

creased from today's typical value of 80 kg down to maybe half its value simply by a redesign of the roller assembly. The effect of such a reduction is visualized in figure 5. Notice that a mass reduction also increases the resonance frequency, proportionally to l/ M^~ .

The problem of poor damping of modes having a node close to the stabilizer location can be much reduced by using two stabilizers at different locations. Reference is made to figure 6 showing the effect of an extra dampener placed between the primary stabilizer and the fast sheave. In this case we have used an even higher linear damping ( Β = 2ζ ) ίθΓ each pair of dampers. Also notice that the second stabilizer is purposely placed below the center location between the fast sheave and the first one. The main advantages of this location are 1) we avoid that the two locations represent neighbor nodes and poor damping at twice the original poor damping frequency and 2) both stabilizers will be exposed to nearly the same motion and force as the first one. Bear in mind that the effect of a stabilizer drops with the distance to the fast sheave.

Options for j ¾^

Below are listed some options for different realizations of dampers that may be used in stabilizers in hoisting systems according to the present invention. Later on a preferred option is described in more

Schematic side view and top view of a hoisting system 1 according to the present invention is shown comprising a fast line 2 with a winch, here only shown as a drum 4, fast sheave 6 and stabi- lizer 8 are shown in figure 7. The use of two identical and perpendicular dampers 10 ensures equal damping in the two transversal directions. The roller assembly 14 is here shown with a double set of twin rollers 16 with V-shaped raceways 18, but other types of roller assemblies are possible, using other numbers (minimum 2) and shapes of rollers. Not shown in the schematic views is the anchoring of the telescopic dampers to the rig structure. The schematic drawings also lack means, typically a rod or wires, for keeping the vertical location of the roller assembly. If only two rollers are used, this guide rod may be substantially parallel to the line, sufficiently stiff and hinged with a uni- versal joint near the fast sheave to allow transversal motions but prevent rotational motions of the roller assembly. The support structure to which the stabilizer and the wire rope guiding sheave are connected, is not shown in the figure.

The dampers 10, here visualized as small telescopic cylinders, can have three levels of control These levels are discussed briefly below.

1) P ssive cjamefr

The term passive damper here means that it has fixed energy dissipation characteristics. A candidate for such damper is a shock absorber used as in the suspension of automotive vehicles. A drawback of such shock absorbers is that they have rather non-linear and asymmetric characteristics. The asymmetry means that the damping force is different for compression speeds than for extensional speed. Another drawback is that vehicle shock absorbers seem to have a speed rating of typically 0.5 m/s, which is substantially lower than what may be needed in some embodiments of the present invention.

Another damper option is a balanced hydraulic cylinder, as shown schematically in figure 8. The damper 10 comprises a standard hydraulic cylinder 20 (hatched inner area representing oil) with hydraulic ports 22 in both ends. But instead of having only one piston rod it has two rods 24, 26 of equal diameters. Only the left one carries the axial load. The air filled cylinder to the right protects the dummy rod and carries the reactive load from the main cylinder. A not shown rubber bellow around the air exposed part of the left rod is recommended for protecting this rod too. The main advantages of this design are 1 ) the inner volume is constant thus creating zero extension force if a common pressure is applied in both chambers and 2) the damping characteristics are the same in both speed directions. When a transversal motion of the fast line makes the piston move, a resulting pressure difference across the piston will make the hydraulic fluid will flow partly through a fixed metering valve through the piston, and partly through an external bypass line of variable restriction.

If simple metering orifices are used, the damping force will be nearly quadratic: F oc -v|v| . This progressive non-linearity is probably far better that the other extreme non-linearity represented a Coulomb friction F oc—sign(v) , because larger transversal line speed in general requires higher damping. If a more linear characteristics are highly desired, it is possible to use more advanced metering mechanisms, for instance, a variant found in some MC shock absorbers where a stack of thin plates are used at the end of the metering channels. When the pressure difference and flow rate increases, the stack is also lifted thus providing increased flow area and less restriction as compared with a constant orifice. Another possibility for linear characteristics of the external bypass line is to use either a single or a group of multiple parallel small cross section flow lines. If the flow in these lines is laminar, the pressure drop will increase proportionally with the flow rate. A drawback of this kind of linear behavior is that the fluid viscosity and thereby also the damping coeffi- cient will be rather sensitive to the temperature.

The term semi-active means that the energy dissipating characteristics can be changed rather rapidly to adjust the damping characteristics dynamically. A possible candidate for such a damper is controllable shock absorbers found in some high end car brands, provided that the speed and force ratings are adequate. Key words are smart fluids and electric control of the damper characteristics. The most common smart fluid is magnetorheological fluids. As the name indicates, its viscous properties can be changed almost instantaneously by a magnetic field. It is mostly the gel strer and thereby the apparent Coulomb friction that can be controlled by the magnetic field. A high band width control of the rheology can, in principal at least, be used to mimic a linear damper action. Because certain embodiments of the present invention may need to handle speed pulses of very short durations, actually down to 15 ms at maximum line speeds, it is questionable whether such a rapid control is feasible. A simpler option is to control the rheology more slowly in response to line tension, line speed and damper fluid temperature.

The balanced cylinder already mentioned as a passive damper option, could also be regarded as a semi-active damper, especially if it includes an external, controllable bypass valve in parallel with a metering orifice through the piston. The damping can be controlled in many ways. One of the simpler options is a stepwise flow resistance control, for instance by opening or closing one or more external bypass lines. A more advance option is to use an external metering valve that can be proportionally controlled to provide a continuous variation of the flow resistance. To allow for thermal expansion and also some small but finite leaks the oil volume should be connected to an external oil reservoir with an optional gas accumulator. This reservoir should preferably be connected to the center of the external bypass flow line to hinder an oscillating accumulator flow.

An alternative to hydraulic dampers is eddy current brakes. Such brakes have fairly linear characteristics, at least for moderate speeds and magnetization levels. It means that the braking force is proportional to the speed and to the stator current producing the magnetic field.

¾ Active damper

Active damper here means an actuator that can handle two-way energy flow, not only energy dissipation. Both hydraulic and electric actuators can be used. Assume that the actuator itself can be represented by an inertia mass, M _c , so that the sum of the stabilizer and controller masses, M_c + M_s , is acted upon by the sum of the line force and the actuator controller force. Assume also that the controller is a general PID speed controller meaning that the controller force at angular frequency ω can be written as

The set speed, v_sel , can either be set to zero, or to the fraction of the low pass filtered lateral speeds at the drum: v_{r lp}l I L and v_{a ip}l I L . Notice that the integrator gain, / , has the dimension of linear stiffness whereas the derivative gain D has the dimension of mass.

The total, effective impedance of the stabilizer, including the mechanical stiffness S = Τ/Ί is simply the sum of the mechanic impedance and the controller induced impedance. That is,

S + I

Z,^ = Z, + Z_c = ( + 5) + + /<y( _s + M_c + D) (25)

A normal PID controller has non-negative values for the factors for / and D , so at a first glance it seems optimal to use a P-controller only, where P is adjusted so that the total damping is approximately equal to the characteristic line impedance. That is,

P « Tm - R (26) However, there is no rule stating that the integrator and derivative gain factors cannot be negative. Without such restrictions it is possible to achieve partially or even total impedance match over a wide frequency range by setting

T

/ - -c,S - -c. (27)

D = -c, (M_s + M_c) (28) where c_s and c, are stiffness and inertia compensation factors, respectively. In practice, full compensation, c_s = 1 and c, = 1 , may not advisable, mainly because of the risk for instabilities. Stability problems can arise if there are time delays on the speed controller, or if the estimates for the mechanical stiffness (or tension) and masses have significant errors. However, partial compensa- tion, up to 80-90% say, should be practically feasible if the time delays in the control loop are sufficiently low. By increasing the damping beyond the characteristic impedance, by a factor of 2 or 3 say, both the stability risk and the damping band width can be improved. These improvements are achieved on the expense of sacrificing a nearly full impedance match and high damping near the stabilizer resonance frequency.

The above controller impedance represents the ratio of telescopic force divided by telescopic speed. In practice, the basic moving element is a rotating motor connected to the linear motion by a gear mechanism. Such a motion converter, which could be a reduction gear combined with a ball screw or a rack and pinion mechanism, can be characterized by a transmission radius, r_c , being the ratio of linear speed to angular motor speed. It can be shown that the impedance for linear, telescopic motion is related to the angular motor axis impedance, Z_m , through

β(&η¾¾¾ί1«-¾¾οη to -oblique d mper m&anftrta

The examples above are worked out for the special case when the dampers are both normal to the line and mutual perpendicular. The results can be generalized to more general orientations of the dampers. Let a_x and /?, denote respective the deviations from pure radial and longitudinal directions for damper 1 having a telescopic impedance Z_c , , and let a₂ , β₂ and Z_{c 2} be the corresponding parameters for damper 2. With longitudinal direction is here meant the direction of line between center or rollers and the pivoting point at the top of the stabilizer bar. This direction may differ slightly from the longitudinal line direction.

It can be shown that the effective damper impedances in respective axial and radial directions can be written as

Z_a = Z_{s a} + (cos a, sin _x)² Z_{c l} + (cos a₂ sin β₂)² Z_{c 2} (30)

Z, = Z_s + (sin a, sin β f Z_{c X} + (sin a₂ sin β₂ f Z_c>2 (31 )

Here Z_{s a} and Z_{s r} represents the mechanical impedances of the roller assembly in axial and ra- dial and directions, respectively. It is easily verified that the two expressions reduce to the first ex- pression of equation (24) when Z_c , - Z_{c 2} , β_λ - β₂ = π/2 and α - -a₂ - π/4 . Note that the latter angles represent the angles in the projected plane normal to the line axis and not in the plane spanned by the two dampers.

Using the angles a_l = -a₂ = π/4 instead of a, = 0 and a₂ = π/2 represents several practical advantages. First, it makes it is easier to anchor both dampers to the rig structure. Second, each damper will get lower peak strokes, speeds and accelerations. The reduction factors for al! these variables are approximately cos( r/4) = 0.71 . Third, it helps to distribute the wear of the telescopic damper almost evenly over its entire stroke. A damper exposed to the radial motion only (a = 0) would have a small working range thereby making seal and rod wear much more concentrated. A similar reduction of the telescopic stroke, speed and acceleration is obtained by a slant angle. Now the reduction factor is ύη(β) . A disadvantage of a small slant angle,

« π/2 , is that the guide rod must handle an increasing longitudinal force equal to E, cot β , F_t being the resulting transversal force. If this rod is fastened to the roller assembly with an offset distance between its neutral axis and the fast line, then the force will also generate a bending moment being proportion- al to the offset and to the longitudinal force. The required rigidity and weight of the guide rod will therefore increase with decreasing slant angle.

The simulation model

A simulation model is developed both for supporting the linear theory above and for being able to calculate the effect a realistic stabilizers having non-linear damping characteristics. The model is based on the convective version of the wave equation (18). The convection, which means that the line is moving longitudinally with a constant translation speed V , is obtained by substituting the partial time derivatives by the so-called material derivatives: d I dt→ d I dt + V■ d I dx . Numerical solutions are obtained by representing the string by typically 200 discrete elements each with a mass AM = mAL , AL being the constant element length. The end nodes, representing the drum and fast sheaves, are modelled with nearly infinite masses to prevent force induced lateral motion. The stabilizer node is also specially treated with a mass AM + M_s and a speed dependent external damping force. The force is, in general, non-linear and represented by the formula

F_d -

. {v] ₊ vl†-^lV v_s (32) where D_Sf. is a generalized damping coefficient representing the damping force at a stabilizer speed of v_s = 1 m/s , ε is a speed exponent and v₀ is a small transition speed (typically 0.01 m/s) included for numerical reasons. This speed exponent equals 1 for linear damping, 0 for Coulomb friction and 2 for a quadratic damping but it can have any value between the two last extremes.

It is useful to explore the damping of the balanced hydraulic cylinder visualized in figure 8 and also being suggested as a working embodiment. Assume that the hydraulic oil is forced to flow through an internal and/or external restriction and that the restriction creates a pressure drop described by the Bernoulli equation AP = v, (33)

where v_m = v_s A_h j ' A_m is the speed of the fluid through the metering restriction, A_h is the hydraulic cylinder area and A_m is the effective cross section area of the metering restrictions. The corresponding damping force is simply

F_d = -A»Ap = - v, (34)

The last equation can be used for finding the metering cross section as a function of the hydraulic area and the chosen damping coefficient /),., :

The simulation model is implemented in Simulink, a powerful simulation tool under the Matlab umbrella. Examples of simulation results are shown in figures 9 and 10, representing the respective cases without any stabilizer and with one stabilizer having two identical dampers mounted as visualized in figure 7. The line parameters are the same as in the numerical example used for the linear analysis. Each damper is a passive hydraulic one as visualized in figure 8. The total mass of the roller assembly and the stabilizer is 40 kg and the quadratic damper coefficient is

D_s2 = 4000 N/(m/s)² = 4.7 ζ/im/s) . Both cases cover a simulation time window of 16 seconds where the drum starts at rest at the beginning of layer 2 and accelerates to a steady rotation speed of approximately Ω « 4 · 2ns^~] = 2A()rpm . Approximately means that the actual speed is purposely reduced by the resonance shift factor 1 - (v/c)² « 0.97 due to the longitudinal line speed of v « i 5 m/s .

The selected speed corresponds to line excitation frequencies being multiples of the cross-over frequency f_xa » 2Ω,/2π = 8.0Hz . This frequency is the 8^th harmonics of the theoretically predicted line resonances, see figure 4. The simulation model confirms that the above shift factor is correct, because resonances are not hit spot on if the frequency shift factor is neglected for the high line speeds.

The simulation results basically confirm what was predicted by linear theory. When the cross-over frequency hits a line resonance, large vibrations build up in the line. The sharp resonances predicted by the linear theory also imply that it takes time to reach steady state vibrations. With reference to figure 9 we see that steady state is not completely reached before the drum spooling enters layer 3 and the excitation direction reverses. This sudden polarity reversal of speed excitation pulses explains why the dynamic line vibrations temporarily reduce when the spooling starts on the 3^rd layer. Simulation with extended time intervals, not shown here, show that the severe line vibrations pick up again beyond 16 s.

Similar simulations are also carried out with one stabilizer (having two perpendicular dampers) placed near the fast-sheave, see figure 10. When comparing with the previous results (without any stabilizer figure 9) we observe that vibrations are now much smaller in amplitude. Additional simu- lation results, not visualized in graphs here, show the following:

1. The vibration reduction is surprisingly robust against changing in changes of the damping characteristics. Only minor changes in the peak fleet angle are observed if the quadratic damping coefficient is increased or reduced by a factor of 2. Moreover, switching from quadratic to linear damping characteristics also seems to have a surprisingly small effect on the peak acceleration, provided that the damping force magnitudes at typical transversal speeds around 1 m/s is roughly the same in both cases.

The simulator confirms the big difference in response predicted by figures 3 and 4 when the stabilizer position is flipped. When the stabilizer is placed a distance 5 m from the drum, high resonance vibrations are seen for drum speeds having a harmonics frequency that matching the predicted resonance at about 10.8 Hz.

When the damping is non-linear, the dynamic line vibrations and axial fleet angles are not independent on the stabilizer orientation a_x even though the two identical stabilizers are identical and perpendicular ( a_x - a₂ = π/2 ). This is because combined motions in the stabilizer directions have different patterns and peak values, as visualized in the speed subplots in figures 1 and 2.

The input speed and acceleration at the drum end of the line are independent of the line dynamics. In contrast, the corresponding variables at the stabilizer are very sensitive to the stabilizer mass and damping characteristics. In the examples above the axial peak speed were reduced by a factor 6, from 3.36 m/s and for the non-stabilized case to 0.56 m/s for the stabilized case. The corresponding figures for the axial acceleration are 29.0g and

3.4*.

The simulation model also provides values for the stabilizer force and for the dissipated power. In the enclosed case the peak force were 1450 N while the peak power is in the order of a few hundred watts. The low pass filtered power is far less, in the order of SOW, which represents only a marginal heating of the hydraulic cylinder.

The stabilizer, when placed relatively close to the fast sheave, has virtually no effect when the fundamental excitation frequency (the double drum rotation frequency) hits a resonance having a node at the stabilizer location. It is therefore recommended to choose an even shorter distance between the fast sheave and the stabilizer so that it is less than half the wave length for the combination of empty block line tension and maximum drum speed. Alternatively, one can use two stabilizers as discussed below.

Simulations with two stabilizers, each with two perpendicular dampers with a relatively high damping ( M_x = M₂ = 40kg ; Β_χ = Β = 2ζ ), mounted at different locations (z_x = -0. IL and z₂ = 2z,/3 ) strongly indicate that the vibration damping can be substantially improved as compared with the single stabilizer solution discussed here. Especially the poor damp- ing of modes having a node at the first location is very much improved. This improvement confirms what was predicted by the impedance spectrum for dual stabilizers visualized as the solid curve in figure 6.

A preferred set of fast line dampers are two balanced hydraulic cylinders as visualized and mounted in figure 7, right and left, respectively. Key parameters for the roller assembly and cylinders are

M_s - 40kg effective stabilizer mass (roller assembly + rods + piston)

/ = 0.10 L Distance between damper and fast sheave (tangent point)

L_d - 0.6m Length of damper (in center position, 50 % extended)

AL_C = 0.2m Maximum stroke

a. = - ₂ = π/4 Azimuth angles (relative to the pure radial direction)

β_{ = β_Ί - π/2 Inclination angles (relative to the line direction)

A_h - 884mm² Hydraulic area (from D_rod = 30mm and D_cyl = 45mm ) ε = 2 speed exponent (quadratic damping)

D_s2 = 4000 N/(m/s)^e damping coefficient

The chosen damping coefficient corresponds to a restriction area of A_m = 8,66mm² , calculated from equation (35) when using a fluid density of p = 870 kg/ m³ .

A typical drum length (width) is X_d - 2.0m so the chosen stroke can easily handle a damper position of maximum / = 0.101 below the fast sheave. This is because the total theoretical stroke for both dampers will be cos · X_d l/L = 0.14m . The suggested cylinder and piston rod diameters correspond to a hydraulic area of A_h - 884mm² , which in turn corresponds to a pressure of Ap ss 5.6 MP a at an estimated maximum damping force of F « 5 AN . This is far below a standard pressure rating Ap ¾ 2 \MPa (or 3000 psi). Standard hydraulic fittings and seals are therefore applicable. The suggested metering area represents a progressive damping. Assuming that the Bernoulli equation for pressure drop applies, that is, Ap∞

where p « 850&£w^-3 is the hydraulic fluid density and v_m ^~ v_d A_d / A_m is the fluid speed through the metering nozzle in the case when the external bypass is closed. The corresponding effective impedance or force to speed ratio then becomes F/v_d = ApA_d /v_d « 0.5 |v_rf | A_d / A_m ² . With the suggested dimensions and a telescopic damper speed of v_d - \ .0ms^'x this ratio becomes F/v_d « 2900A¾m^_1 . This value is about 4.5 times the characteristic impedance of a 1.75" fastline exposed to a tension of 50 kN. if the external bypass loop is opened, the total bypass area is up to tripled and the effective impedance at the same speed is reduced by a factor from 1 to 1/9, the latter representing a damping less than the characteristic impedance. With reference to the theory above it is anticipated that these dampers will reduce the fastline vibrations substantially. If the external restriction is adjusted manually, the dampers can be regarded as passive. In contrast, if the restriction is varied automatically in response to the drum speed and line tension for achieving optimal damping characteristics over a wide range of operational conditions, then the dampers can be classified as semi-active.

The roller assembly, which connects the dampers to the fast line, is not discussed in details here because there are many designs that will work almost equally well. The suggested variant in figure 5 can probably be designed to a mass of less than 30 kg resulting in a total mass of one dampener less than 40 kg. To minimize noise from the relatively small rollers to the dampers and rig structure it is advisable to use elastomer (rubber) sleeves in the couplings, similar to those used in many vehicle shock absorbers. Such coupling will act as mechanical low pass filter reducing transfer of high frequency vibration components and transmitting mainly the low frequency components. Noise reduction and smoother roll action can also be obtained by substituting parts of the V-profile of the rollers by a U-profile having a curvature radius matching the maximum radius of the line.

It should also be noted that the hoisting system according to the invention may utilize a variety of different stabilizers and that the theory outlined above and the insight derived therefrom do not limit the invention to any specific type of stabilizer. In addition to the embodiment discussed in detail above, the stabilizer may also take the form of a sleeve through which the fast line may be running, where the sleeve may be strung to the derrick (support structure) in a variety of different ways in order to obtain the desired damping effect. The advantage of using a sleeve instead of the roller- type embodiment described above, is that is reduces the number moving parts and thus also the risk of falling objects. The sleeve may be mounted around the fast line with guide bushings instead of rollers discussed above.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design many alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. Use of the verb "comprise" and its conju- gations does not exclude the presence of elements or steps other than those stated in a claim. The article "a" or "an" preceding an element does not exclude the presence of a plurality of such elements.

The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

Claims

C l a i m s

1. Hosting system (1) for a drilling rig, the hoisting system (1) comprising:

- a wire rope (2);

- a winch (4) for pulling in and letting out said wire rope (2),

- a support structure, such as a derrick;

- a wire rope guiding sheave (6) connected to the support structure, said wire rope guiding sheave (6) being provided between said winch and a load suspension member along the wire rope (2); and

- a first stabilizer (8) for damping lateral vibrations of the wire rope (2) between said winch (4) and said wire rope guiding sheave , c h a r a c t e r i s e d i n that said first stabilizer (8) is provided closer to the wire rope guiding sheave (6) than to the winch (4) along the wire rope (2).

2. Hoisting system (1) according to claim 1 , wherein said first stabilizer (8) comprises one or more dampers (10) with first end(s) connected to said support structure, said first stabilizer (8) further comprising one or more pairs of guide rollers (16) common each of said one or more dampers and connected to a second end of said one or more pairs of dampers (10), said wire rope (2) being adapted to run between the rollers (16) of each of said one or more pairs of guide rollers.

3. Hoisting system (1) according to claim 2, wherein said first stabilizer (8) comprises a pair of dampers (10), wherein longitudinal axes of the two dampers are substantially perpendicularly oriented relative to each other.

4. Hoisting system (1) according to claim 3, wherein each of said substantially perpendicularly oriented dampers (10) are provided at ± JT/4 radians, respectively, relative to the radial direction of a winch drum of the winch (4) in a plane normal to the length axis of the wire rope.

5. Hoisting system (1) according to any of the claims 2-4, wherein said one or more dampers (10) are passive dampers having substantially non-adjustable damping characteristics.

6. Hoisting system (1) according to any of the claim 2-4, wherein said one or more dampers (10) are semi-active dampers having controllable energy-dissipating characteristics.

7. Hoisting system (1) according to any of the claims 2-4, wherein said one or more dampers (10) are active dampers having controllable damping characteristics and controllable active reaction.

8. Hoisting system (1) according to any of the preceding claims, wherein said hoisting system further comprises a second stabilizer, said second stabilizer preferably being provided between said wire rope guiding sheave (6) and said first stabilizer (8) along said wire rope (2), and even more preferably between said wire rope guiding (6) sheave and said first stabilizer (8) closer to said first stabilizer (8) than to said wire rope guiding sheave (6) along the wire rope (2).

9. Hoisting system (1 ) according to any of the preceding claims, where the said first, and potentially second, stabilizer (8) has a moving mass in the order 12 percent or less of the fast line mass, preferably 6 percent or less..

10. Hoisting system (1 ) according to any of the preceding claims, wherein said first stabilizer (8) is provided at in the order of 1/10 of the distance from the wire rope guiding sheave (2) to the winch (4).

11. Drilling rig comprising a hoisting system (1 ) according to claim 1.

12. Drilling rig according to claim 11 , said drilling rig further comprising:

- means for measuring the rotational speed of a winch drum of said winch (4);

- means for measuring the tension in said wire (2) rope between said winch drum and said wire rope guiding sheave (6);

- a control unit adapted to receive said measured rotational speed and said measured tension.

13. Drilling rig according to claim 12, wherein said control unit is adapted to use said measured winch drum rotational speed and said wire rope tension as input in a model for calculating optimized characteristics of said first and/or second stabilizer, said control unit, if said first and/or second stabilizer are of semi-active or active type, further being adapted to adjust the motion characteristics of said first and/or second damper.

14. Method for damping lateral vibrations in a wire rope (2) by means of hoisting system (1) according to claim 1 , wherein the method comprises the following steps:

- measuring the rotational speed of a winch drum of the winch (4);

- measuring the tension of the wire rope (2) between said winch (4) and said wire rope guiding sheave (6); and

- calculating optimized motion characteristics, for reducing said lateral vibrations, of one or more dampers (10) of said first stabilizer (8) based on said measured rotational speed and tension.

15. Method according to claim 14, wherein the method, if said dampers (10) are of a semi- active or active type, further comprises: - adjusting the motion characteristics of said dampers (10) based on said calculated, optimized motion characteristics.