Enhancement of flow rates through porous media

[00002] This is a Continuation-in-Part of patent application Ser. No. 09/828,177, filed Apr. 9, 2001, (now U.S. Pat. No. 6,405,797, issue date Jun. 18, 2002), which is a Divisional of patent application Ser. No. 09/046,762, filed Mar. 24, 1998, (now U.S. Pat. No. 6,241,019, issued Jun. 5, 2001), which carries a priority date of Mar. 24, 1997 from GB-97/06044.

[00003] This invention relates to the dynamic enhancement of fluid flow rates in a porous medium, using pressure and strain pulsing. The invention relates to devices and methods designed to explicitly enhance the flow rate of fluids (liquids or gases) and mixtures of fluids and solids (e.g. oil and sand particles) in porous media by means of application of pressure pulsing or strain pulsing to the region of flow. The pressure pulsing is applied to the liquid phase of a porous medium through periodic cycling of liquid volumes by mechanical, hydraulic, or pneumatic devices at one or more points. Strain pulsing can similarly be applied through mechanical or electromechanical excitation. The two processes are intimately linked in that a pressure pulse generates a strain pulse, and vice-versa. Dynamic enhancement of fluid flow rate can be applied to the following technologies:

• • [00004] Flow of liquids or liquid-solid mixtures to wellbores in petroleum or water extraction processes from porous media.
  • [00005] Flow of liquids or liquid-solid mixtures in porous media to wells, sumps or other pressure sinks during cleaning of contaminated shallow aquifers comprised of sand, gravel, or fractured rock.
  • [00006] Flow of liquids or liquid-solid mixtures in contained or natural porous media beds used for chemical engineering reaction processes, filtration, refining, cleaning, or other processes where liquids or liquid-solid mixtures are flowing from one point to another under the effect of a pressure or gravity-induced gradient.

[00029] In the context of this specification, a porous medium is a natural or man-made material comprising a solid matrix and an interconnected pore (or fracture) system within the matrix. The pores are open to each other and can contain a fluid, and fluid pressure can be transmitted and fluid flow can take place through the pores. Examples of natural materials include gravels, sands and clays; sandstones, limestones and other sedimentary rocks; and fractured rocks including fractured sedimentary rocks which have both fractures and pores through which fluids may flow. Examples of man-made porous media include filtration beds of natural or artificial granular materials or manufactured solid porous materials, as well as beds of catalysts used to accelerate reactions between fluid phases or fluid-solid phases during refining, chemical synthesis, or other processes. Structures such as tailings dikes, dams, fluid recharge or filtration beds, and so on, can be regarded as porous media.

[00030] The porosity of a porous medium is the ratio of the volume of open space in the pores to the total volume of the medium Systems of practical interest in the present context have porosities that lie in the range 5% to 60%.

[00031] The porosity (pore, fractures, and channels) is filled with fluids, which may be gases or liquids or a combination of the two. Liquids can be oil, water (with dissolved constituents), or man-made liquids such as gasoline, chlorinated bi-phenyls, polymers, and non-aqueous phase liquids deliberately or accidentally introduced into the porous medium. Gases may be natural hydrocarbons, air, carbon dioxide, or man-made gaseous products introduced deliberately or accidentally into the porous medium.

[00032] All porous media are characterized by a permeability. Permeability is an average measure of the geometry of the pores, pore throats, and other properties which describes the flow rate of fluids through the medium under the effect of a pressure gradient or a gravity force induced because of differences in density among fluid phases or solid-fluid phases.

[00033] Pressure pulsing is a deliberate variation of the fluid pressure in the porous medium through the injection of fluid, withdrawal of fluid, or a combination of alternating periods of injection and withdrawal. The pressure pulsing may be regular or irregular (periodic or aperiodic), continuous or episodic, and it may be applied at the point of withdrawal or at other points in the region of the porous medium affected by the flow process.

[00034] Strain pulsing is a deliberate variation of the strain at a point or local region in the porous medium by applying changes in strain through a device which vibrates, oscillates, or which expands and contracts in volume. The strain pulsing may be regular or irregular, continuous or episodic, and it may be applied at the point of withdrawal or at other points in the region of the porous medium affected by the flow process.

[00035] Dilational and shear pulses are the two basic types of excitation. In a dilational pulse, the perturbation is isotropic (equal in all directions) at the point of application, and may be termed a volumetric pulse. Pressure pulsing is dominantly a dilational perturbation. The dilational perturbation moves out in all directions approximately equally and is subject to scattering phenomena. In a shear pulse, a relative lateral excitation is applied so that the energy imparted to the porous medium is dominated by shear motion, such as occurs when slip occurs along a plane. Shear perturbation is highly anisotropic, and the distribution of energy depends on the orientation of the perturbing source. Shear perturbations can therefore in principle be focused so that more energy propagates in one direction than in another. Strain pulsing can be anisotropic or isotropic, depending on the nature of the excitation source.

[00036] Flow takes place in a porous medium through generating a pressure gradient in the mobile (moveable) phases by creating spatial differences in fluid pressures. Reducing or increasing the pressure at a number of points may generate this by the withdrawal or injection of fluids. It may also be generated through the force of gravity acting upon fluids of different density, such as oil, formation water, gas or air, injected non-aqueous phase liquids and other fluids. In a system where the solid particles are partly free to move, density differences between solids and fluids may also lead to gravity-induced flow.

[00037] In a porous medium containing two or more non-miscible fluids (oil and water for example), the wetting phase is that gas or liquid which, because of surface tension and wettability effects, is in contact with the majority of the solid material. It forms the pendular fluid contacts between grains in a granular porous medium, and coats the walls of flow channels (FIG. 1). The non-wetting phase is that gas or liquid which lies in the interstices and channels and is separated from the solid material by a film of the wetting phase fluid. In FIG. 1, the mineral grains 1 are coated with a wetting liquid 2, while a non-wetting liquid 3 occupies the rest of the pore space. The pore throat dimension 4, averaged through the medium, is important in dictating the velocity at which liquid can pass through the pores 5 of the medium.

[00038] The non-wetting phase 3 might be continuous or discontinuous. If it is continuous, then an interconnected and uninterrupted path of that liquid exists in the medium. If it is discontinuous, the non-wetting phase may exist as isolated droplets or regions, which are nowhere in direct physical contact with other regions of the same phase.

[heading-00039] 2 Evidence of Dynamic Enhancement of Fluid Flow

[00040] There exist in the public literature observations of increased flow rates in oil wells and water wells during and after dynamic excitation from earthquakes or other events which can create sufficient strain in the medium to affect the porosity, and the through-flow velocity of the liquid, even in a minuscule manner.

[00041] In the systems as described herein, periodic or irregular pressure pulsing in a flowing system under a pressure gradient increases the flow rate of the mobile phase toward the extraction point.

[00042] Field observations confirm that a porosity perturbation applied to a petroleum well enhances flow to the well for some time thereafter by increasing the mobility of the fluid phase. In the case of a petroleum well producing fluid and sand, a general increase in the mobility of the complex solid-liquid-gas flowing phase takes place. The perturbation in these cases may also be a single sharp pressure pulse applied at the production well.

[00043] Theoretical developments and field observations show that fluid flow rate to a producing water well or petroleum well is enhanced if the liquid-flow-borne solids are allowed to enter the wellbore in an unimpeded manner. This is analogous to a porosity diffusion process in that a porosity change occurs as the solid phase is produced along with the liquids. This porosity change slowly propagates out from the production point into the porous medium through a diffusive mechanism, and is accompanied by changes in the pressure and pressure gradient with time and location around the wellbore. In the oil industry, the process of allowing the sand to flow unimpeded is called cold production, cold flow, or sand production.

[00044] In general, the flow enhancement accompanying any porosity diffusion process takes place in a system with a pressure gradient, and the processes preferentially increase flow rates of the mobile, non-wetting phase if more than one fluid is present as a continuous phase.

[00045] One feature of the invention lies in the recognition that dynamic excitation through application of a pressure pulse, a strain pulse, or a series of pulses anywhere in the flowing porous medium can enhance the flow rate. Fluid rate enhancement occurs at the exit points of a given system (wellbore, reaction bed, and pipeline), that are also the points of low pressure in the medium. Furthermore, we have recognised that the fluid flow enhancement can be theoretically predicted and analysed, measured in the laboratory, and physically explained.

[00046] In addition to the porosity diffusion effect and the enhancement in flow rate that it generates, dynamic excitation has several other beneficial effects on production performance of wells. The dynamic excitation may be induced as a pressure pulse or a strain pulse, generated by a pulsating or vibrating source. Excitation may be periodic or aperiodic, continuous or episodic, and applied in the stratum or at the surface, provided that sufficient porosity diffusion amplitude is transmitted to the region of interest.

[00047] The permeability of a conventional producing well can be impaired by the migration and consequent accumulation in the near-wellbore environment of fine-grained solid particles, which can pass through the pore throat constrictions in the porous medium. When, as described herein, the porous medium is being dynamically excited the tendency for these particles to bridge and block porosity is substantially reduced, thus allowing the well to maintain flowing conditions with a minimum of impairment.

[00048] Particularly in viscous heavy oils but also in some conventional oils, certain liquids (asphaltenes in general) can be precipitated as small size solid particles when the liquid encounters the lower pressures near the wellbore. These particles can accumulate in the pore throats, impairing the permeability of the system and reducing the flow rate to the producing well. Dynamic excitation, as described herein, provides cyclic strain energy aimed at mitigating the tendency for blockages of these precipitants, maintaining the well in a superior flowing condition.

[00049] Finally, under conditions where the granular particles of the porous medium are allowed to flow along with the fluids (as in sand production), the flowing particles may bridge together near the wellbore, forming a stable sand arch, and stop the solids flow. This condition leads to a massive deterioration in the fluid productivity of the well. Dynamic excitation, as described herein, provides a perturbation energy, which tends to destabilize these arches because of the small cyclic strains induced at the contacts between sand grains.

[heading-00050] 3 Experimental Verification

[00051] FIG. 2 shows an experimental set-up 20 to demonstrate the physical principle of dynamic enhancement of fluid flow. The cylindrical device 23 contains a dense sand pack 24, which is under an applied stress of 1.5 MPa. The sand pack is flushed through with paraffin oil (or any other wetting phase) to coat the grains as a continuous wetting phase. Then, glycerin (or other non-miscible liquid) is allowed to flow through the sand and form a continuous non-wetting phase that is immiscible with the wetting phase. The fluid exit port 25 allows production under the action of a pressure gradient maintained constant by keeping a reservoir 26 of the mobile non-wetting phase liquid 27 at an elevation higher than the device 23. Exit port 25 has a screen 28 between the port 25 and the sand pack 24 for experiments where the sand is not allowed to flow; however, for experiments where the sand is permitted to flow, the screen is removed.

[00052] The flow experiment is allowed to reach a condition of steady-state exit port flow rate Q. Once this condition is reached, a dynamic perturbation is applied to the system by one of two methods: a small strain pulse is applied through a transducer embedded in the sand 24; or, a periodic pulse is applied to the upstream part of the device by perturbing the flexible flow lines manually or automatically (at point 29). The varying excitation is indicated by the symbol in the circle. Pressure transducers (P1, P2, and P3) are electronic devices designed to monitor any changes in pressure in the system induced by the dynamic excitation. The sand pack 24 is maintained in compression by hydraulic pistons 35.

[00053] The strain pulse is applied through a small acoustic transducer linked to an oscilloscope and signal generator 30. The acoustic transducer (not shown in FIG. 2) is embedded in the sand 24 during the assembly of the experiment. It has a diameter of 15 mm and is encased in latex to seal it from the fluid and to provide good coupling with the sand pack. Being of such small size with respect to the cell, it does not impede the flow of liquids through the experimental apparatus. The frequency of the sonic pulse was varied from 10 Hz to 60 Hz during the excitation period in the experiments. The period of excitation is indicated in FIG. 3 as pulsing-started 32 to pulsing-stopped 34. In between periods of excitation, no pulsing takes place, but flow is allowed to continue; this is necessary to evaluate flow enhancement through contrasting periods of excitation and periods of no excitation, in the same apparatus without other changes on the pressure head or flow properties.

[00054] The pressure pulsing is applied by manually squeezing the upstream flexible tube 29 connecting the fluid reservoir 26 to the top of the flow apparatus. This manual squeezing is applied at a frequency of 0.5 to 2 Hz continuously during the excitation period.

[00055] FIG. 3 demonstrates quantitatively the change in the flow rate from the experimental device. The lower line 36 is the steady flow at a hydraulic head of 0.25 meters (the top of the fluid in the reservoir was maintained at an elevation of 0.25 meters above the entry port). This line 36 is to demonstrate that without pressure or strain pulsing, no flow enhancement takes place. The upper line 37 is the demonstration of enhancement. In this case, the fluid reservoir was maintained 0.5 meters above the fluid entrance port, and one may note that the slope of the non-pulsed portions of the line 38 is almost exactly twice the slope of the lower line 36. This is in accordance with the conventional view of flow through porous media: a doubling of the hydraulic head without pulsing leads to a doubling of the flow rate.

[00056] The slope of the upper line without pressure pulsing or strain pulsing (38) is approx 2.67 cm³/min. With pressure pulsing or strain pulsing, the flow rate increases (37) to approx 5.7 cm³/min, an enhancement factor of about 2.15. Various experiments conducted with different excitation frequencies and excitation times showed flow rate enhancement factors of from 1.5 to 2.2, demonstrating that the porosity diffusion process increases the flow rate of the mobile phase under conditions of continuous pressure or strain excitation.

[00057] This enhancement is also observed in a set of experiments where the sand is allowed to move from the exit port (screen is removed). Experiments where sand was allowed to exit are intended to simulate the behavior of wells producing heavy oil or other liquids by the process of sand production, discussed below in more detail. Results similar to those shown in FIG. 3 are obtained if the sand in the specimen is allowed to exit. Flow rate enhancement ratios of 2.0 to 2.5 are typically obtained. Typical results are shown in FIG. 4. The only difference in experimental set-up between this figure and the previous one is that now the sand is allowed to flow out with the fluids at the exit port.

[00058] In the sand+liquid flow experiments (screen 28 removed), it was observed that after some time the sand spontaneously stops exiting because of the formation of a stable sand arch behind the exit port 25. This blockage causes the fluid exit rate to drop to a negligible value, <0.2 cm³/min, indicating that the sand grain arch is impeding the flow of liquids. The pulsing and the strain perturbations overcame this blockage. The results therefore indicate that not only is there a basic flow rate enhancement, but also that the natural tendency of sand to create blockages can be overcome by pressure or strain pulsing, and if such blockages exist, they can be de-stabilized by pulsing. Clearly, this has substantial positive implications on maintaining free fluid and sand flow to a well producing sand and liquids.

[00059] FIG. 5 shows the pressure response from the three pressure transducers in FIG. 2 (P1, P2, P3) when the device is subjected to a series of continuous pressure pulses applied by manually squeezing the inflow hose at point 29. As mentioned earlier, the actual magnitude of this pressure pulse is less than 0.2 kPa, and it has no effect on the average pressure head applied to the sand pack. With continued pressure pulsing, however, the actual fluid pressure in the specimen begins to rise (the curves swing upward); this is the effect of the porosity diffusion process being built up through the continuous excitation. When the pulsing is stopped, the pressure enhancement begins to decay slowly back to its original values, but the flow rate at the exit port drops to its initial values within 2-5 seconds. This suggests that fluid flow enhancement requires continuous excitation. FIG. 5 shows that the pressure build up is less the farther away from the excitation source because the pressure build up attenuates as the porosity diffusion wave is transmitted through the system.

[00060] FIG. 6 shows details of the experimental set up where a small-embedded acoustic transducer A1 (or several small transducers) is providing dynamic excitation. This excitation is of extremely small amplitude, yet it has the same effect as the pressure pulsing: it alters slightly the pressure in the fluid phase, and also changes the stresses between the grains, which builds up the pressure in a way similar to FIG. 5. This also is a porosity diffusion process because the acoustic excitation is a small-amplitude strain wave, which leads to small perturbations in the porosity of the porous medium. Experimental data show that this process also leads to a fluid flow rate enhancement of the same order of magnitude as the pressure pulsing, and the enhancement effect can also be predicted and analysed theoretically.

[heading-00061] 4 The Physical Effect in Cold Production Wells

[00062] The proposed technology has wide applicability to a number of conditions and cases. However, we believe that it has particular value in the petroleum industry. Therefore, we describe in detail one production process, Cold Production (CP), which will be substantially aided by the application of dynamic pressure or strain pulsing. This detailed presentation is in no way meant to exclude any of the other possible production practices for conventional oil, heavy oil, or other fluids present in porous media. This example was chosen because it has two major aspects of the beneficial effect of dynamic excitation through pressure or strain pulsing: the effect of increasing basic flow rate, and the effect of breaking down the stable sand arches that form and tend to block oil flow.

[heading-00063] 4.1 Cold Production Mechanisms

[00064] It is best to have a clear understanding of the production mechanisms involved in the oil rate enhancement observed during Cold Production (CP) in order to understand how pressure or strain pulsing can enhance flow rates and prevent blockages through the formation of sand arches.

[00065] First, movement of the solid matrix (sand) directly increases the velocity of the fluid (oil+water+gas). Thus, sand movement increases flow velocity, enhancing production. This can be seen in FIG. 4, where the initial slope of the flow line 39 when solids and fluids are both allowed to flow is greater than for the case of no solids 37, even under the same hydraulic head.

[00066] Second, sand extraction creates a more permeable zone around the wellbore through dilation of the sand matrix from an average of perhaps 30% porosity to a porosity of 35-38% porosity. This zone grows in mean radius as more sand is produced (some wells produce in excess of 1200 cu m of sand in their lifetime). If the growth of this zone is stopped or impeded by sand blockages, flow rates will be lower. If stable sand arches form near the well perforations, the flow rates may drop to a small fraction of their values when the sand is free to flow. If these sand arches are continuously destabilized by dynamic excitation so that they cannot form in a stable manner, oil flow is not only more continuous, but it occurs at a greater rate.

[00067] Third, dissolved gas (mainly CH4) in the heavy oil exsolves gradually in response to a pressure drop. Bubble nucleation and gas exsolution is retarded in time because of low gas diffusivty in viscous oil. The gas also tends to remain as a separate bubble phase during flow toward the wellbore, and bubbles expand as the pressure drops toward the production site, giving an internal drive mechanism referred to as foamy-flow. It is believed that the foamy flow mechanism aids solids extraction and enhances fluid flow rate. The high viscosity of the oil retards gas exsolution during flow, and bubble mobility in the pores and throats is retarded by interfacial tensions. This alters permeability and enhances development of small-scale tensile stresses, which help destabilize the sand.

[00068] Fourth, asphaltene precipitation and pore throat blocking by clays or fine-grained minerals are reduced during CP because of continuous solids movement, which liberates pore-blocking materials. Regular pulsing of pressure or strain will greatly reduce the frequency of pore throat blockages, which may arise.

[00069] Oil production in CP wells can, exceptionally, be as high as 20-25 cu.m/day, although 4-10 cu.m/day is more typical. After prolonged CP, done conventionally, rates as low as 1 or 2 cu.m/day can be accepted providing that initial rates were sufficient (e.g. >5 cu.m/day) for a long enough period (e.g. 2 years) to warrant well drilling and field development. However, the systematic application of pressure or strain pulsing is expected to extend the productive life of a well, and will also increase the production rate of the well on a daily basis.

[00070] CP mechanisms depend on continued sand movement, which allows foamy oil mechanism to operate efficiently, and which allows continued growth of a disturbed, dilated, partly liquefied region around the well.

[heading-00071] 4.2 When Cold Production Stops

[00072] Some wells in Alberta have produced oil and sand stably for over 11 years, with sand flow being successfully re-established after workovers, or even during production. However, some wells are extremely difficult to maintain on stable sand production. Generally, a failure to sustain solids flow is directly related to a major drop in oil production. Therefore, re-establishment of sanding would have positive economic consequences in increased oil rates or prolonged production periods. This re-establishment can be a consequence of a continuous destabilizing of the formation, unblocking perforations, or otherwise destroying any stable structures, which may have been generated in the sand. Dynamic excitation, as described herein, is aimed at achieving these goals.

[00073] Stable sand structures are desirable, for good CP. These include: stable perforation sand arches which greatly retard fluid flow into the well; re-compaction of the sand in the near-wellbore environment; collapse and blockage of flow channels within the strata; or perhaps generation of some form of natural gravel-pack created by a natural settling around the wellbore of the coarser grains in the formation. Changes of fluid saturation leading to increases in capillary cohesion have been suggested as a common blocking mechanism. This idea suggests that gas evolution leads to increasing gas saturation near the wellbore until a continuous gas phase exists, with an apparent cohesion increase in the sand.

[00074] Little is known in detail about the actual blocking mechanisms because of difficulties in exploring the wellbore region and difficulties in laboratory simulation, and therefore there is some difficulty over a method of evaluation and implementing ameliorative measures. What methods are used have been arrived at empirically and developed through practice. To our knowledge, no one uses pressure or strain pulsing of a continuous nature during continued production.

[00075] Workovers have been used to perturb the formation and re-establish sand ingress. The conventional methods used vary from surge and swab operations to much more aggressive approaches such as Chemfrac (TM), involving igniting a rocket propellant charge to blow materials out of the perforations, as well as to shock the formation and perturb the sand. Considering the rise time and the fluid velocity, this method is probably the most effective in unlocking perforations plugged with sand and small gravel particles. However, none of these methods are continuous in nature during the production of the well.

[00076] Mechanical sand bailers on wirelines are conventionally used to clean the well of sand before replacing a worn pump. The bailer is dropped repeatedly into the sand until filled, and then withdrawn at a relatively rapid rate. This has a vibrational effect on the near wellbore area, and a swabbing effect during withdrawal. Often, after bailing, sand has flowed back into the well through the perforations, and cases have been reported of six to eight days of bailing, removing as much as 1-3 cu.m of sand; that is, 10-15 times the amount that was in the wellbore in the first place. Bailing is relatively successful in re-establishing sanding, but extensive periods of bailing are clearly to be avoided if better alternatives exist.

[00077] Injection of various chemical formulations to break capillary effects is relatively common, as is injection of several cubic meters of heated oil. These methods are thought to break any apparent (capillary) cohesion in the sand, and the outward flow is thought to reopen some perforations that may have become blocked.

[00078] Thus, although many conventional wells produce sand freely, blockages occur. The method of strain or pressure pulsing, as described herein, through the process of porosity diffusion, can provide long-term continuous production at an enhanced flow rate by activating the ambient stress field dynamically. This process can destroy small-scale stable sand arches and keep pore throats open. Blocking materials such as asphaltenes and clay particles are much less likely to plug pore throats under conditions of dynamic pressure or strain excitation.

[heading-00079] 5 Field Configurations

[00080] FIGS. 7 and 7a show an example of how dynamic enhancement through pressure pulsing can be implemented in the field. A pressure pulsing system is installed in the central well 40 of a porous stratum containing oil and water. Perforations in the steel casing 42 of the well 43 allow full and unhindered pressure communication between the liquid in the wellbore and the liquids in the pores and fractures of the porous medium. The well is completely liquid-filled between the pulsing device and the perforations, and is maintained in that condition.

[00081] A number of adjacent wells (H1, H2, H3, and H4) are producing fluids and therefore have a well pressure that is less than the excitation well 40. In other words, the pressure gradient in the porous medium is directed by the induced pressure differences so that fluid flow is toward the producing wells. FIG. 7b shows a typical pressure decline curve between the excitation well and the producing wells. The distance d between the well 40 and the producing wells 43 is dictated by the physical properties of the medium (compressibility, permeability, fluid viscosity, porosity, thickness, fluid saturation), and must be determined through calculations and field experience for individual cases. The pattern shown, or any other suitable pattern of producing wells and excitation wells, may be repeated to give the necessary spatial coverage of a producing field.

[00082] In the field, the amplitude, frequency, and waveform of the dynamic excitation can be varied to find the optimum values required to maximize the dynamic enhancement effect. Because porous media have certain characteristic frequencies at which energy dissipation is minimal, analysis, laboratory experimentation, and empirical field optimization methods (based on outflow rates at the producing wells and other monitoring approaches, discussed below) might be required to find the best set of operating parameters which maximize the dynamic flow rate enhancement. Monitoring approaches for optimizations are discussed later.

[00083] FIG. 8 shows another possible configuration for implementation of pressure or strain pulsing to enhance fluid flow to wells. For illustration purposes, suppose that a vertical well 45 is completed with a number of short-radius laterals 46, each of which is considered a horizontal well. Fluid is to be withdrawn through the well 45 with the horizontal drains. A number of excitation wells 47 are emplaced above the horizontal laterals, and pressure pulsing or strain pulsing is applied in these wells through excitation devices 48.

[00084] In both cases pulsing can be generated either through a downhole or a surface pressure pulsing which can be activated by mechanical, hydraulic or pneumatic means.

[heading-00085] 6 Pressure and Strain Pulse Devices for Oil Exploitation

[00086] FIG. 9 shows one example of a pressure pulsing device that causes a periodic pressure excitation at a controllable frequency and amplitude. The pressure pulsing can be varied in frequency (number of pulses over a time interval), in amplitude (magnitude of the pressure pulse), and in waveform (the shape of the pressure pulse). The pulsing is governed from the surface through an appropriately designed electronic or mechanical control system. The major elements of the diagram are:

• • [00087] a) A wellbore 50, having a casing 52, embedded in cement 53, perforated into the target formation 54.
  • [00088] b) A piston pump barrel 56 which, when mechanically actuated, generates a pressure pulse.
  • [00089] c) A one-way valve 57 to allow entry of fluid into the zone below the piston pump on the upstroke of the piston.
  • [00090] d) An actuating device, in this drawing represented as a rod 58 to surface within the production tubing 59 that is isolated from the casing annulus with a packer 60. This driving mechanism can be varied in frequency and stroke length (volume).

[00091] The driving mechanism for the piston pump 56 in FIG. 9 is a surface-driven reciprocal or rotary mechanical drive that creates an up-and-down motion of the piston 56. Alternatively, the driving mechanism can be an electromechanical device above the piston pump driven by electrical power. Alternatively, a surface pressure impulse can be applied through the tubing. In this case, the piston pump may be replaced by a flutter valve top-hole or bottom-hole assembly which opens and closes to create pressure surges which enter the formation 54 through the perforations, but does not affect the annulus pressure because of the packer 60.

[00092] The piston 56 may contain the one-way valve 57 to allow intake of fluid on the upstroke, and expelling the incremental fluid on the down stroke, generating the pressure pulse. Alternatively, the fluid valve 57 can be closed, and a periodic pressure impulse generated with a closed system.

[00093] As shown in FIGS. 10a, 10b, and 10c, a single well 62 is producing fluids through perforations in the steel casing 63 because the pressure in the well is maintained at a value lower than the fluid pressure in the far-field, generating a pressure gradient which drives fluids (or a fluid-solid mixture) to the wellbore 62. The examples show both an inclined well and a vertical well integrated with progressive cavity pump system for purposes of illustration only. Operational descriptions will focus on a rotating elliptical mass, but it is understood that the principles apply to other pulse-like sources of strain energy.

[00094] FIG. 10a shows a typical down-hole assembly for the application of a periodic mechanical strain to the casing in the producing formation, and the cemented casing serves as a rigid coupling system that transmits the periodic straining to the formation. The major elements of the diagram are:

• • [00095] a) A cased 63 cemented wellbore perforated into the target formation 64 with tubing assembly and other peripheral devices.
  • [00096] b) A fluid pump 65 to withdraw fluids and sand from the wellbore 62.
  • [00097] c) Housings and devices that couple the fluid pump 65 to the tubing and if desired to the well casing 63, through a rigid packer (not shown).
  • [00098] d) A system of rods 67 connecting the fluid pump to the drive mechanism.
  • [00099] e) A drive mechanism to give rotary action to the fluid pump and eccentric mass 68.
  • [00100] f) An eccentric mass 68 which is mechanically linked to the fluid pump 65 (FIG. 10c).

[00101] Installed in the wellbore is a mechanical or electromechanical device that applies vibrational energy to the casing through rotation of an eccentric mass or through volumetric straining. The device is fixed to the exterior casing 63 through conventional means, using a packer with steel contacting pads (slips) or other means whereby the vibrational energy is efficiently transmitted to the steel casing with a minimum of energy losses. A schematic cross-section of a rotating elliptical mass is given in FIG. 10c. The central square hole is stabbed by a square rod on the bottom of the power rods 67, which are rotated from the surface. As the rods rotate and thereby also activate the fluid pump 65, the eccentric mass is rotated at the same angular velocity, or else the velocity may be less or greater if a mechanical gearing device is included.

[00102] The rotation of the eccentric mass 68 creates an imbalance of force, which causes the casing 63 to apply a rotational strain to the surrounding porous medium through which the casing penetrates. The rotational strain generates an outward moving porosity diffusion wave that perturbs the liquid in the porous medium, causing an accompanying pressure pulse in the liquid. The energy thereby applied to the liquid (and entrained mobile solids) in the porous medium leads to an enhancement of liquid flow into the wellbore, irrespective of the direction of propagation of the porosity perturbation. Furthermore, the strain energy thereby applied reduces or eliminates tendencies for the material pore throats or fractures to become blocked by fine-grained particles, precipitants, or through the formation of stable granular arches. The fluid produced is removed from the wellbore through the pump 65, which in this example sits above the elliptical rotating mass, but the order of the devices may be altered. Both the pump and the rotating mass may be mechanically driven, electrically driven, or one may be mechanical and the other electrical.

[00103] FIG. 10b shows a typical down-hole assembly for the application of a periodic mechanical strain to the casing in the producing formation, and the cemented casing serves as a rigid coupling system that transmits the periodic straining to the liquid in the formation. The major elements of the diagram are:

• • [00104] a) A cased cemented wellbore 69 perforated into the target formation 70 with tubing assembly 72 and other peripheral devices.
  • [00105] b) A progressive cavity (PC) pump 73 to withdraw fluids and sand from the wellbore 69.
  • [00106] c) Housings and devices that couple the stator of the PC pump 73 to the tubing 72 and if desired to the well casing, through a rigid packer, not shown.
  • [00107] d) A system of rods 74 connecting the PC pump to a drive mechanism 75.
  • [00108] e) A drive mechanism 75 to give rotary action to the PC pump and eccentric mass 76,
  • [00109] f) An eccentric mass 76 which is mechanically linked to the PC pump.

[00110] The driving mechanism for the PC pump 73 in FIG. 10b is a surface-driven rotary mechanical drive that creates a variable frequency rotation of the rods 74, rotor, and the eccentric mass 76. Alternatively, the driving mechanism for the bottom-hole assembly can be an electromechanical device located above or below the rotor, and driven by electrical power.

[00111] The device that applies a large rotational strain to the casing, is an eccentric mass driver 76, which is rigidly coupled to the rotor of the pump 73. The strain is a circular impulse triggered by rotation of a mass that is located off the centre of rotation of the PC pump assembly, and it may be located above or below the rotor. To transmit the strains effective to the well casing, it is necessary that the rotating eccentric mass be rigidly coupled to the casing. This is achieved through a packer seating assembly (not shown) either below or above the PC pump, but close to the eccentric mass, so that the rotary impulse is efficiently transmitted.

[00112] FIG. 11 shows an approach to transmit periodic mechanical energy down the tubing assembly in a cased well through application of mechanical excitation at the surface. These strains are transmitted to the bottom of the well, where they may be converted to a pressure pulse, or mechanically linked to the casing to transmit mechanical strains to the liquid in the formation. The major elements of the diagram are:

• • [00113] (a) A cased cemented wellbore 78 perforated into the target formation (not shown).
  • [00114] (b) A tubing and rod assembly.
  • [00115] (c) A drive head 79 that rotates the rod to provide motive power to the bottom-hole fluid pump (not shown) which may be a progressive cavity pump or a reciprocating pump.
  • [00116] (d) A packer device 80 to allow the polished section of the tubing 82 to undergo a periodic vertical movement independent of the casing or the rods.
  • [00117] (e) A driving mechanism 83 of variable frequency and stroke that imparts a vertical periodic motion to the tubing 82, separate and distinct from the pump drive-head 79.
  • [00118] (f) A set of reaction springs 84 and a flange 85 on the wellhead to act in unison with the tubing drive mechanism 83 to give the periodic vertical movement.
  • [00119] (g) Housings and devices that isolate yet allow the movement of the tubing and rods to allow production from the well while tubing excitation is active.

[00120] The example shown in FIG. 11 is a rotating motor actuating the tubing through an eccentric cam, with counter-stroke reaction provided by a set of springs. A variety of other driving mechanisms can be used, including a direct mechanical linkage of a reciprocating device to the tubing (perhaps eliminating the springs).

[heading-00121] 7 Pressure or Strain Pulsing in a Reaction Bed

[00122] A reaction bed (FIG. 12) of granular or porous material 86 is used to foster chemical interaction by introducing two fluid species (liquid-liquid or liquid-gas). The pore-and-throat structure similar to that in FIG. 1 of the porous medium helps break up the two fluids into intermingled phases with a large surface contact area, which accelerates the reaction process. The solid phase may, for example, be an inert material such as silica particles, or it may be a bed of particles of catalyst or of ceramic particles coated by a catalyst. In the case of a catalyst, the use of a porous bed gives a high surface contact area between the catalyst and the reacting phases. The flow through the system is achieved either through downward gravitational flow, or through a difference in the fluid pressure between the input and exit ports. The flow in this case may be in any direction, but always in the direction of the induced pressure gradient. In the example shown, flow is from top to bottom. To increase the efficiency of the process, the flow rate of the fluids through the reaction bed should be maximized.

[00123] Fluid rate flow is accomplished through the application of pressure pulses on the reaction bed by pulsing the pressure in the liquid inflow lines (S1, S2) or exit lines (SS), or by applying pressure pulses through a port (T) or ports in liquid (pressure) communication with permeants. Alternatively, vibrational strain energy can be applied either externally or internally (U1-U4) through the use of mechanical devices or electro-mechanical transducers. The symbol inside the small circles indicates that pulsating pressure or strain is being applied at these points.

[00124] In these cases, porosity diffusion processes and the coupled pressure-strain responses create the necessary flow enhancement effect.

[heading-00125] 8 Pressure or Strain Pulsing to Facilitate Aquifer Remediation

[00126] We give the example of cleaning of a potable water aquifer that has been contaminated by a non-wetting phase, which has permeated the pores and exists as a continuous liquid phase. Using strategies, which, for example, may be of similar configurations to those in FIGS. 7 and 8, and devices presented in FIGS. 8 to 10b, pump-out wells are configured to give the best areal coverage of the contaminated water reservoir. Furthermore, excitation leading to fluid flow enhancement through porosity diffusion effects at these shallow depths can be implemented as well at the surface, through the use of harmonic oscillators, for example (not shown).

[00127] The aquifer clean-up proceeds by continued pumping and can also be enhanced by the input of water or other suitable liquid or solid/liquid mixtures at the excitation wells, or at other wells installed specifically for this purpose. The key aspects in this case are the continued excitation, the continued provision of a source of liquid to account for the voidage generated by pumping the wells, and the maintenance of a pressure gradient in the aquifer that maintains flow to the low pressure production (clean-out) wells.

[heading-00128] 9 Monitoring and Optimization in the Field

[00129] Periodic straining or pulsing can enhance the flow rate in a porous medium. The excitation gives rise to dynamic porosity diffusion effects. Optimization of the excitation process involves determining the most effective frequency, amplitude, and waveform to be applied. Control of the excitation is applied through a controller and a power source, with an oscilloscope or other read-out device to examine the characteristics of the excitation.

[00130] In order to optimize the process, it is necessary to monitor both the excitation effects and the production rate. This is achieved through monitoring production rates using flow meters or tank gauges, and through monitoring the transmission of the excitation within the reservoir. The important excitation factors to monitor are the nature of the excitation and the nature of the waves transmitted through the reservoir, and these data are collected at a data acquisition system connected to a computer. The parameters of importance in the reservoir are the pressure and the wave trains. The pressure is monitored at a number of points through pressure ports in observation wells and excitation wells, and the wave train is monitored using geophones, accelerometers, or other suitable devices placed in observation wells, excitation wells, or behind the casing in production wells.

[00131] In order to optimize the process, the data streams are taken to a central computer where the data are plotted and correlated. Then, the parameters are optimized to allow maximization of the production rate, subject of course to the limitations of the equipment used for the excitation.

[heading-00132] 10 Criteria for Site Selection

[00133] The preferred framework for field implementation of dynamic enhancement is outlined below. It is designed to answer a number of basic requirements to facilitate proper site selection, which should increase the probability of successful implementation and oil recovery.

[heading-00134] 10.1 Reservoir Porosity

[00135] The effect of vibrational enhancement is relative to the current parameters, which make economic recoverability viable. For example, porosity simply determines the amount of oil in the reservoir. It does not, in theory, play a direct role in the effectiveness of the process until large porosity values are obtained. It is suggested that for maximum effectiveness the bounds of porosity range from 18% to 35%. At porosity levels above 35% the effects dynamic enhancement becomes less cumulative, diminishing with further increases. Below about 18% (i.e oil shale), enhancement by pulsing would not be expected to occur.

[heading-00136] 10.2 Minimum and Maximum Porous Media Thickness

[00137] An aim of dynamic enhancement through application of pressure or strain pulses, as described herein, is to propagate a slow moving wave in three-dimensional space. This may be in an oil reservoir or in a system comprised of a natural or artificial porous media. For optimum operation, the preferred constraints on propagation of a continuous or episodic pressure or strains in the systems described previously are as follows:

• • [00138] a) For oil reservoirs and aquifers a minimum thickness of 3.0 meters to a maximum of 50.0 meters.
  • [00139] b) For contained reaction beds, a minimum thickness of 0.20 meters to a maximum of 1.0 meters.
    10.3 Caprock

[00141] Caprock, the geomaterial that overlies an oil reservoir or aquifer, serves two purposes. First, it prevents the pressure or strain pulse from upward propagation beyond the parameters outlined in Section 10.2, and it prevents upward flow of fluid. A caprock may consist of shale, dolostone, salt (or other evaporites), very dense clays, tight limestones, and so on. The key element for a caprock in the case of pressure or strain pulse propagation is that it be of extremely low permeability (e.g salt), or have very low permeability (e.g shales, dolostone, and very dense clays). It is important to note that the propagation of the pressure or strain pulse propagates through the liquid in the porous medium. It is the elastic properties of the matrix and the mobility and compressibility conditions of the fluid, which will ultimately determine the viability of the process. If the matrix is weak, or brittle, the matrix might tend to crack and consolidate under the action of pulses that have enough energy to create the dynamic enhancement of liquid flow rate as described herein. In that case, the invention would be contra-indicated. The caprock conditions are of a secondary nature but are listed here for completeness.

[heading-00142] 10.4 Permeability

[00143] The ratio of viscosity to permeability defines the mobility of the oil in a reservoir, a contaminant in an aquifer, of fluid in a reaction bed. The range of permeability for aquifers and bed reactors preferably should be on the order of 1000 sq.cm (gravel) to 0.01 sq.cm (silt). For light oil and heavy oil reservoirs the dynamic enhancement process is viable at a range from 10⁻¹¹ cm2 to 10⁻¹³ cm2.

[heading-00144] 10.5 Viscosity

[00145] The magnitude of the diffusion constant and the scale of the interaction determine the speed of the pressure or strain pulse. The diffusion constant is directly proportional to permeability divided by viscosity. From our calculations of the speed of propagation of a pressure or strain pulse without the advantage of large tectonic stresses in the earth or large hydraulically induced stresses (i.e. bed reactors) we place the cutoff at 30 API gravity. When the earth's tectonic stresses can be used as an energy source both grain slippage and fluid flow will effect the propagation speed of the pressure or strain pulse. In this case, and in the case of high hydraulic stresses, the cutoff to oils can be as low as 10 API gravity.

[heading-00146] 11 Estimation of Enhanced Fluid Production from a Pulse Series

[00147] It has sometimes been observed, after an earthquake, that the flow rate of liquid through a porous medium has been significantly improved, at least for a time. This has led to techniques and proposals for subjecting the porous structure to artificial seismic perturbations. However, the technique of applying pulses to the liquid in the porous medium is quite different from the technique of imparting seismic perturbations to the medium itself, being much less disruptive (and less costly). Besides, although seismic operations might open up the pores, it might happen instead that the medium consolidates and closes the pores; the system as described herein is aimed rather at pulsing the liquid (and any grains that might be entrained in the liquid) relative to the solid matrix, not at pulsing or shaking the solid matrix itself.

[00148] A quantitative estimation of the cumulative enhancement of fluid production, which is observed in porous media subjected to a periodic impulse, depends on the geometric disposition between the pulse generator and the production wellbore. Such a quantitative estimate can be achieved for an arbitrary geometry through numerical calculations based on the pressure pulse and a porosity diffusion model for earthquake sources or explosive perturbations. Those perturbations produce irreversible changes in porous media, i.e. fracture, dilatancy and compaction. Any impulse-triggered decrease of porosity leads to effective compaction, and this can squeeze an additional amount of fluid from the porous medium. From a physical point of view this mechanism is clear, and such a mechanism is known to lead to excess pore pressures and sand liquefaction during strong earthquakes.

[00149] In contrast to the irreversible compaction arising from single strong perturbations, the invention is aimed at providing reversible strains arising from continuous weak perturbations. Each perturbation (e.g. tapping or short-term cyclic straining) is assumed to be of an elastic nature which does not produce any residual, irreversible deformation, but does cause a periodic perturbation in the porosity of the system through compression and relaxation.

[00150] In the aftermath of an impact, a porous medium relaxes to the equilibrium state in a diffusional manner because the relaxation process involves flow of the viscous saturating fluid with respect to the porous skeleton. If we apply another perturbation before the proceeding one fully decays, while withdrawing the produced fluid through a port, a cumulative, synergetic effect can be achieved. A quantitative estimation of this effect for specific cases involving non-symmetric dispositions of the perturbation source and the wellbore requires extensive analytical and computer model calculations based on numerical methods. This approach, however, tends to obscure the physical logic on which the model is based.

[heading-00151] 12 Further Considerations

[00152] An aim of the invention is to apply pressure pulses and strain pulses to a liquid in natural and man-made porous media to enhance the flow rate of the mobile fluid phases and to diminish the probability of flow-rate impairment through the internal bridging of particles. The approach has been verified theoretically, in the laboratory, and through empirical observations in field situations in the petroleum industry and for water wells. A key element is the concept of porosity waves and attendant pressure pulses, which travel through the medium by diffusional processes. To our knowledge, this phenomenon has not been previously identified in such media and considered for the purposes of fluid flow rate enhancement. Applications are envisioned particularly but not exclusively for the petroleum industry and the chemical processing industry. Also, in reservoirs contaminated by non-aqueous phase, non-wetting liquids, implementation of pressure pulsing and other means of generating porosity diffusion enhanced flow is expected to accelerate clean-up operations, and make them more effective.

[00153] The techniques as described herein should be distinguished from fluidized bed technology, in which a granular material is pulsed at such an energy level that the whole solid matrix is in a state of heaving motion. In the present case, the intention is that the solid matrix does not move, but rather that the pulses pass through the liquid while the solid matrix remains substantially stationary.

[00154] Liquid flowing through a porous medium has a flow rate, which depends on the impressed pressure differential. Within the porous medium, the velocity of the liquid, as caused by that impressed pressure differential, will vary from pore to pore, but the velocity may be averaged as a volumetric flow rate over the whole treatment volume.

[00155] Considering a pore P: if the porosity of pore P should decrease, i.e if the pore should close up, the velocity of liquid passing through that pore would go down, for a given impressed pressure differential. The porosity might go down if, for example, a grain of sand might become snagged in the pore.

[00156] The pressure pulses spread through the liquid, as a wave-front, with a wave velocity. The wave front velocity (and magnitude) will not be the same at every pore in the treatment volume. The velocity of propagation of the wave-front may be averaged over the treatment volume.

[00157] In a real porous medium, the average velocity of propagation of the wave-front will be much faster than the average flow-through velocity of the liquid. Similarly, at each pore, the velocity of propagation of the wave-front will be much faster than the velocity of the liquid travelling passing through the pore.

[00158] The pressure pulse, as it passes through a pore, causes a surge in the liquid present in the pore. As the wavefront passes, the pressure differential across the pore increases, and so the through-flow velocity of the liquid in the pore momentarily speeds up (assuming the wave-front is travelling in the same direction through the pore as the liquid). Afterwards, the pressure differential across the pore drops back, as the wavefront passes, and the liquid in the pore slows down and reverts back to the background velocity of the liquid through the pore.

[00159] If the wavefront were travelling against the liquid travel velocity, the pulse would cause the velocity of the liquid in the pore to drop momentarily, then gradually speed up again to the background velocity, as the pulse passes. In some cases, the velocity of the flow of liquid in the pore might even reverse (and back flush the pore) momentarily.

[00160] It is the sudden changes in the through-velocity of the liquid in the pore that prevents grains settling in the pore, whether the pulses cause a momentary speeding up of the liquid in the pore, or a momentary slowing down (or even reversal) of the liquid in the pore.

[00161] Thus, the pores are kept open by the surges. The sudden change in velocity of the liquid dislodges or flushes away grains that might be snagged in the pores, and prevents grains from snagging in the pores. It may be noted that an actual reversal of the flow velocity of the liquid can be especially effective, by back-flushing the pores clear. By sweeping or flushing the pores clean, the flow rate of liquid through the treatment medium can be increased; or, at the least, the rate at which the pores become clogged can be slowed. An even more beneficial ratcheting effect also can be engineered. The pulses have a specific wave form, which includes a gradual rise in pressure, followed by a gradual fall in pressure. See FIG. 5. (The wave form at pore P might not be the same as the wave form as created by the pulse generating means.) Insofar as this pressure pulse gives rise to a change in the pressure differential across the pore, the velocity of the liquid in the pore undergoes a change that follows a similar waveform.

[00162] If the pulses are infrequent, the next junior) pulse reaches the pore P after the earlier (senior) pulse has died away, and so each pulse of pressure has an independent, i.e non-cumulative, effect on the through-velocity of the liquid passing through the pore. This condition is illustrated in FIG. 14a. However, if the pulses are more frequent, the junior pulse might reach the liquid in the pore before the senior pulse is finished. That is to say: the senior-surge in the flow rate of the liquid through the pore is still present when the junior-surge in the flow rate arrives. The senior-surge in liquid flow rate is caused by the pressure differential imposed by the senior pulse, and the junior-surge in liquid flow rate is caused by the pressure differential imposed by the junior pulse.

[00163] With the arrival of the next pulse after that, the velocity of the flow of liquid in the pore is given a further incremental increase, and so on. This condition is shown in FIG. 14b.

[00164] The effect is repeated in all the other pores, and thus the effect is manifested as an increase in the overall flow rate of the liquid through the treatment volume of the porous medium. It has been found that the velocity of the flow of liquid through the treatment volume can be increased asymptotically to an upper limit 93 (FIG. 14b), which is considerably faster than the background flow rate 94 arising simply from the differential pressure imposed on the treatment volume without pulsing. That is to say: the flow rate is increased by the pulsing as if a larger pressure differential had been imposed, or as if the porosity had been increased.

[00165] Thus, not only does the pulsing as described herein tend to keep the pores clear as the changes in flow velocity flush the pores, but also the pulsing, if done at the right frequency, can increase the actual flow rate of the liquid through the treatment volume.

[00166] The frequency of the pulses should be rapid enough that a junior pulse arrives at the pore before the senior pulse has died away. On the other hand, the frequency of the pulses should not be too rapid. Too high a frequency might set up resonances in the solid matrix material, and cause the material to undergo an amplitude of movement that might cause damage. Also, the higher the frequency, the more it becomes difficult to get enough energy into each pulse to actually cause a significant pressure surge in the liquid, per pulse.

[00167] The engineer should carry out tests at the treatment site, in which the overall through-flow rate is measured for different frequencies of pulsing. The frequency should be increased (starting from about 1 Hz) until a frequency is reached beyond which no further increase in through-flow rate is achieved. Typically, that happens when the frequency of pulsing is in the range 1 Hz to 10 Hz.

[00168] The magnitude or energy of the pulses is important. If the energy of the pulses is too high, the solid matrix material can be damaged. That is to say, the matrix material should not be shaken so vigorously as to cause some consolidation of the material, which would thereby lose some porosity and permeability. The energy should be high enough, though, to make the momentary change in the velocity of the liquid passing through the pores significant.

[00169] It will be understood that, in many cases, the liquid flowing through the pores will have some sand or other solid grains entrained in the flow. The sand grains of course come from the solid material making up the matrix. The movement of the sand grains, entrained in the moving liquid, should be distinguished from consolidation of the matrix, which involves a settling movement of the matrix material.

[00170] The direction of the pulses is important. In some case, for example if the pulses are generated actually in the extraction well (as in FIG. 10b, for example) the wave-front of pulses propagates in the direction away from the extraction well. In that case, the change in pressure differential across the pore, due to the pulse, acts to create a momentary velocity which opposes the velocity of the liquid through the pore towards the extraction well, due to the imposed background pressure differential. It might be possible in that case, by adjusting the frequency of the pulses, actually to reduce the flow rate of the liquid through the pores, i.e to impose on the liquid such a cumulative effect upon the velocity or flow-rate that the pulse-created flow-rate opposes the background flow-rate. Of course, significantly dropping the flow-rate would run counter to the aims of the invention, and the engineer should see to it, when operating a system in which the wave-front velocity is in the opposite direction to the liquid flow-rate velocity, that the frequency of pulsing stays out of the range in which flow of the liquid towards the extraction-well might be seriously attenuated. The ratcheting of flow velocity as shown in FIG. 14b only applies when the pulses are travelling in the same direction as the liquid.

[00171] By correctly setting the pulsing frequency, the pulsing can be used to prevent clogging of the pores, by flushing the pores and resisting the possible snagging of grains in the pores, whether the wave-front velocity is with or against the liquid extraction velocity.

[00172] One of the dangers of using a separate excitation well to generate the pulses is the possibility of inadvertently establishing a preferred pathway through the porous material, from the excitation well to the extraction well. If that happens, the well would be finished, in that now the liquid being pulled out of the extraction well is simply the liquid being fed in at the excitation well.

[00173] A separate excitation well is useful in that the engineer will find it easier to create the type of pulses that will make a significant difference to the flow rate of the liquid if he not only provides a separate excitation well, so that the direction of the pulses reinforces the flow-rate of liquid towards the extraction well, but also if he injects a (small) charge of liquid into the excitation well with each pulse. Injecting a charge of liquid at each pulse delays the drop-off or fall-back of flow-rate velocity after the pulse passes, which makes it easier to achieve the ratcheting of the pulses that can create a significant improvement in flow rate.

[00174] However, as mentioned, when using an excitation well, the engineer must make sure he does not kill the production well. It is recognised that the pulses can be made to travel considerable distances through the porous medium; sufficiently far indeed that the excitation well can be placed far enough away that the danger of killing the well becomes negligible, and yet the pulses can be made to penetrate large distances into the porous medium.

[00175] It is emphasised that the pulses are pulses of pressure passing through the liquid; the pulses do not require the solid matrix material to move. (Of course, if the liquid pressure changes, a pedant might argue that the solid matrix must undergo distortions corresponding to the change in pressure, if only very slightly. But the invention is concerned with real practical effects, and the pulses as described herein can, as a matter of substance, be generated, and can perform the useful function as described, even if the solid matrix notionally did not move at all.)

[00176] In the case where the pulses are generated as pressure pulses, the pulses are generated by creating motion directly in the liquid; in the case where the pulses are generated as strain pulses, the pulse is first applied to a local region of the solid matrix material, and only indirectly thereby to the liquid. In that case, the solid matrix material undergoes, or might undergo, a measurable strain in launching the pulse into the liquid. However, such a strain would be very localised, as to the distance of penetration of the strain into the porous medium, whereas the pulse that such strain creates in the liquid would then penetrate much further into the porous medium, through the liquid.

[00177] The system shown in FIG. 15 is installed in a borehole 120, having a metal casing 123. The casing is perforated, at 124, to provide an aperture connecting the inside of the casing with the surrounding ground.

[00178] A tube 125 is installed by being passed down inside the borehole, from the ground surface. Packers 126, above and below the perforations 124, seal the tube 125 with respect to the casing 123. The packers are inflatable, from the surface, through packer inflation pipes 127, after the tube has been installed to its correct working depth in the borehole. (In an alternative, the packers are already installed in the borehole, and the tube is stabbed down into and through the packers.)

[00179] A ram 128, comprising a cylinder 129 and a piston 130, is located inside the tube 125, and is sealed into the tube by a ram seal assembly 132. The piston 130 is driven downwards by means of a pneumatic actuator. Air under pressure is fed down from the surface through an actuation conduit 134. A combined actuator and exhaust valve unit 135 receives the compressed air, and controls the flow of air into the actuator cylinder 136. The pressure of the air in the actuation conduit 134 is higher than the pressure of the liquid in the ram cylinder 129, whereby the piston 130 descends when the actuator valve is open.

[00180] The descending piston 130 drives liquid from the ram cylinder 129 into a ram-space 137 inside the tube 125. The tube has a through-hole 138, to permit liquid to pass from inside the tube to outside the tube, which the liquid passes through, and thence into the packer-space 139. The packer-space 139 includes the perforations 124, through which the liquid is forced out into the ground material surrounding the borehole.

[00181] The ram cylinder 129, the ram-space 137, and the packer-space 139, together form a unitary hollow chamber in which the pressurised liquid is contained. The walls that form this hollow chamber are indicated by the dotted lines 140 as shown in FIG. 15A.

[00182] The hollow chamber indicated by 140 is of variable volume, in that as the piston 130 descends, the volume of the chamber becomes smaller. As may be understood from FIG. 15A, the walls that define the hollow chamber of variable volume include portions of the ram components, portions of the tube 125, the through-hole 138, portions of the packers 126, and portions of the well casing 123 that are contiguous with the perforations 124. These and associated components may be arranged differently in different cases, but in each case in which the invention is applied, the components will define a hollow chamber of variable volume, being the chamber which contains and seals the liquid being pressurised, prior to the liquid passing out through the aperture.

[00183] After a volume of liquid, V-out, has been driven out of the perforations, the valve unit 135 is then operated to close off the actuation conduit 134 and to open the exhaust conduit 142. The pressure in the actuation cylinder 136, and the pressure in the ram-space 137, and the packer-space 139, therefore drops. When the pressure in the ram-space falls below the pressure in the inlet conduit 143, the inlet check-valve 145 opens, and admits liquid at the inlet pressure. This drives the piston 130 upwards, the air in the actuation cylinder 136 being exhausted through the exhaust conduit 142. The volume of liquid admitted through the inlet valve 145 is V-in liters.

[00184] The chamber 140 of variable volume has a maximum volume when the piston 130 is against its upper stop 146, and has a minimum volume when the piston is against the lower stop 147. In a typical case where the invention is being used for the remediation of an under-producing oil-well, for example, the structure of the ram may be such that the difference between the maximum volume V-max and the minimum volume V-min, i.e the maximum swept volume, is several tens of liters, or more. (The structure of the ram 128 limits the maximum volume that can be swept by the piston, per stroke; but even if the piston can sweep through a large volume, it may be preferred not to use the full stroke of the piston, every cycle.)

[00185] It will be understood that, in carrying out the remediation operation, the condition that is being sought is a condition that may be termed slosh-pulsing. In slosh-pulsing, the liquid in the ground around the borehole forms a coherent body. Thus, when a volume of liquid is discharged under pressure out of the borehole, the whole of the coherent body of liquid moves outwards, away from the borehole, as a single, unitary, coherent body. When the pressure is relaxed, the coherent body shrinks back, still as a coherent body. Every drop of liquid that makes up the coherent body undergoes an outwards velocity, followed by an inwards velocity, every cycle.

[00186] The slosh-pulsing system is very effective in homogenising the porous ground around the borehole, largely because the coherent body can increase in size and extent gradually and progressively. Thus, at first, only the few liters that lie close to the aperture move coherently, i.e as one with the ram piston. But once a drop of liquid in the ground has been picked up by the coherent body, now it, too, starts to move out/in in response to the down/up cycles of the ram.

[00187] Gradually, as the movements of that drop become consolidated with the movements of the rest of the coherent body, so that drop becomes now firmly attached to the coherent body. So, next, a more remote drop of liquid, which was formerly completely uninfluenced by the pulsing movements emanating from the borehole, now also in its turn is picked up by the cyclic back and forth movements. As soon as the slightest back and forth movement has been induced in the remote drop of liquid, that movement tends to become consolidated, and that drop becomes assimilated into the coherent body. In this way, the coherent body can become very large, and can extend many meters—sometimes many tens of meters—away from the borehole. The back and forth movement of the liquid, as mentioned, opens and clears the pore spaces, and keeps them clear.

[00188] The back and forth sloshing movement of the liquid tends to flush the pores clear of debris. Although the pores themselves may become physically dilated, for a moment, when the liquid in the pores is under a higher pressure, this dilation is temporary, in that the pores settle back to their normal size when the pressure is relaxed. Thus, the porosity of the ground material basically does not change; what changes is the permeability of the ground material, as the pores are flushed and cleared by the back and forth reversing movement of the liquid.

[00189] It is noted that one of the reasons the well is depleted is that the pores around the borehole may have become blocked, primarily because the value-liquid being extracted from the well has been drawn into the well only in one direction, whereby debris has become lodged in the pores. This debris cannot be dislodged very effectively simply by injecting a liquid outwards from the borehole. One problem is that such injected liquid tends to form fingers outwards along the paths of least resistance. The slosh-pulsing, on the other hand, not only tends to clear the pores, but tends not to leave any areas of the ground in which the pores have not been cleared. That is to say, slosh-pulsing tends not only to maximise, but tends also to equalise, or homogenise, the permeability of the ground around the borehole.

[00190] There is a limit to the frequency at which slosh-pulsing can take place. Slosh pulsing involves a volumetric flow of liquid, first with a bulk outwards velocity, and then with a bulk inwards velocity, as a result of a volume of pressurised liquid being injected into the ground, from the borehole. The velocity at which the effect of the injected volume spreads through the coherent volume depends on many factors, including viscosity of the liquid, pore size, etc. Also, the magnitude of the injected volume (V-out) per cycle, and its pressure, are factors that must be properly controlled if the procedure is to be efficacious.

[00191] Slosh-pulsing requires that the volume injected per cycle be significant. The frequency with which a coherent body around a borehole can be made to undergo cyclic sloshing back and forth is very low when the liquid is say a heavy oil, and may be as low as no more than e.g one pulse in several minutes. When the liquid is water, and the ground is quite porous, the max frequency at which slosh-pulsing can be induced may be one pulse per second, or even, sometimes, up to about ten pulses per second. It should be noted that the frequency at which pulsing can be induced might change as the pore-spaces become clear, and as the coherent body becomes larger.

[00192] Generally, the designer will wish to induce slosh-pulsing at as fast a rate as can be supported, to speed up the remedial effects of the procedure. The engineer must be careful not to pulse too rapidly, as that achieves nothing. The effects of too-rapid pulses cannot spread out into the surrounding ground; the analogy may be made of a pendulum, in which the bob of the pendulum would be unaffected if the support of the pendulum were to be displaced and then replaced very rapidly; only if the support is moved slowly can the movement be transmitted to the bob.

[00193] In order for slosh-pulsing to be induceable in respect of a particular borehole, the porous ground around the borehole, containing the liquid, must be able to achieve a condition that may be termed resilient saturation. In this condition, the effect of discharging a volume of liquid into the ground, under pressure, and of then releasing the pressure, is that the residual liquid pressure remaining in the ground outside the borehole, after the pressure in the borehole has been relaxed, is, at least for a short time, higher than the pressure in the borehole. Only if this pressure condition is present can the liquid flow back into the borehole, i.e be sucked back into the borehole by the pressure differential between the ground and the borehole. Porous liquid-filled ground in the resilient saturation condition may be regarded as springy; not only does the springy ground receive the volume that is injected under pressure, but, when the pressure is relaxed, the springy ground drives the injected volume (or at least some of it) back into the borehole.

[00194] It has been found that it is possible to achieve the condition of resilient saturation, and then to induce slosh pulsing, with very many types and porosities of ground material, and with different liquids. The apparatus shown in FIG. 15 can be used to create the required conditions with a minimum of sensors and control equipment, and with a minimum of skill and attention on the part of the operators of the apparatus (given that the operators are well-trained for borehole work).

[00195] At first, when remediation is just starting, and the ground is unsaturated, slosh-pulsing cannot (yet) be induced, and liquid may be simply poured into the ground. The ram may be cycled at this time, but basically the inlet valve 145 stays open all the time, even when the ram piston is descending. But after a time, the rate at which more liquid can be injected into the ground slows, and now the inlet valve 145 does close off, when the ram piston 130 is descending.

[00196] However, at this approaching-saturation state, the pressure in the ground around the borehole drops immediately (and the inlet check-valve 145 opens immediately) the ram piston start back up. Thus, still, no liquid is sucked back into the borehole, and slosh-pulsing has not yet started.

[00197] Then, eventually, a state is reached in which liquid does flow back into the borehole, each cycle, as the ram piston starts to ascend, and the opening of the inlet check-valve is delayed until the piston has risen some way back to the top of its stroke. Thus, if, say, ten liters are injected out of the borehole, two liters might be sucked back into the borehole before the pressure drops enough for the check-valve to open. After that, eight liters is admitted into the chamber, from the inlet conduit, as the cycle is completed. As mentioned, as soon as any volume of liquid is sucked back into the borehole, from the ground, per cycle, the condition of slosh-pulsing has now commenced.

[00198] In order for the apparatus to be able to produce slosh-pulsing, the swept volume of the ram, i.e the difference between V-max and V-min, must be large enough to cause bulk volumetric sloshing movement of the coherent body of liquid around the borehole. That being so, the difference between V-max and V-min, in most cases, must be at least several tens, if not hundreds, of liters. However, as far as the minimum possible volume is concerned, it is stated that the minimum difference between V-max and V-min, if slosh-pulsing is to be contemplated as being possible at all, is more than one liter, per cycle. (It may be noted that where, in the prior art, vibrations have been imparted to liquids in boreholes at acoustic frequencies, the volume of liquid that can be said to physically move, per cycle, is only a tiny fraction of one liter.)

[00199] In practice, an apparatus in which the difference between V-max and V-min, though more than one liter, is less than about ten liters would be of only very limited usefulness, and a general-purpose apparatus of the kind as described herein preferably should be able to deliver at least ten liters of liquid, under pressure, into the ground, by ram action, per cycle. Also, the apparatus should be capable of receiving that same volume back from the ground, per cycle, if slosh-pulsing is to be carried out to its maximum effect.

[00200] It is important that the chamber of variable volume (i.e the chamber that is defined by the dotted lines 140 in FIG. 15A) not be too large. The volume of the chamber would be too large if the difference, in liters, between V-max and V-min was only a small proportion of the number of liters of the overall volume of the chamber.

[00201] For example, the chamber would be too large if the ram apparatus were placed at the surface, and the aperture 124 were at the bottom of a deep borehole. In that case, the equivalent of the ram-space 137 would be very large. The effect of actuating the ram at the surface would then have to be transmitted through the large volume of liquid. The movement of the ram would simply be lost in the compressibility of the liquid, viscous effects, and the like.

[00202] As shown in FIG. 15, the volume of the ram-space 137 and of the packer-space 139, combined, is less than the swept volume of the ram 128, and this small volume of the chamber is possible if the ram is located very close to the aperture 124. But where the ram is further from the aperture, now the volume of the ram-space 137 and of the packer-space 139, combined, might then be larger than the swept volume of the ram. The larger the portion of the volume of the chamber other than the swept volume, the more disadvantageous. However, the apparatus might still be functional, though the ram-space and packer-space are somewhat greater than the swept volume of the ram. By way of an upper limit, for the movement of liquid through the aperture 124 to be responsive, crisply and surely, to the movement of the ram, the ram-space 137 and the packer-space 139 combined should be no more than about fifty times the swept volume.

[00203] The apparatus as here described is intended for injecting liquids into the ground. However, it is contemplated that a similar apparatus be used for injecting a gas into the ground, if circumstances so require. In that case, the need for the ram-space and the packer-space to be small is very much more critical than when the fluid to be injected is a liquid, and in that case, the volume of the chamber other than the swept volume should be just as small as possible. With a gas, the upper limit for the volume of the ram-space and packer-space together is when that volume equals the swept volume (V-out−V-in).

[00204] Slosh-pulsing of a liquid into the ground around a borehole may continue for a long time after the slosh-pulsing first commenced. The remediation of the well continues until, ideally, the whole of the volume injected per cycle is recovered, i.e sucked back into the borehole, per cycle. In that condition, the inlet valve opens only just as the piston reaches the top of its stroke. In practice, this final stage is hardly ever achieved, because, at the remote extremes of the coherent body, liquid can dissipate away without being consolidated into the coherent body, so there is always at least some slight volume to be made up, from the inlet conduit, per cycle.

[00205] It may be regarded that the efficacy of the remediation is measured by how much liquid has been injected, in total, over the whole period of remediation, into the ground. The better the remediation, the greater and the more homogeneous the porosity of the ground, and the larger the volume of liquid that has been injected overall during the remediation.

[00206] It should be noted that, without slosh-pulsing, injection of liquid into the borehole would have stopped at the above-mentioned approaching-saturation stage, where suck-back was just starting to become possible. Injection would have stopped because it would appear (if the apparatus did not enable slosh-pulsing) that the ground was indeed saturated. With slosh-pulsing, it has been found that the ground can be made to accept additional large quantities of liquid, beyond that apparently-saturated point.

[00207] It can take periods that are measured in weeks, or even months, after the ground first appears to be saturated, for the point in the piston up-stroke at which the check-valve opens, gradually to occur later and later in the cycle, until finally the check valve does not open until the ram piston is (almost) at the top of its stroke. In other cases, this final fully-saturated condition is reached after just a few days of slosh-pulsing. Once no more liquid can be injected, remediation is complete, and the apparatus can be removed from the borehole.

[00208] Where slosh-pulsing has reached the point that no more liquid can be injected per cycle, that means that the coherent body of liquid is sloshing back and forth, each cycle, and the reach of the coherent body has extended to the maximum possible limit. Even then, continued back and forth sloshing might serve to clean out and homogenise the pore-spaces still further, whereby after a period of slosh-pulsing during which no more liquid is being injected, it then might become possible to inject some more. It also might be possible that the coherent body of liquid could be made larger, and/or more coherent, if the slosh-pulsing is done at a different frequency or (volumetric) amplitude.

[00209] Thus, the engineer should measure what may be termed the liquid-disappearance-rate, being the rate at which liquid passes down the actuation conduit 134, once slosh-pulsing has started, and should experiment with frequencies and amplitudes, noting whether this or that change causes the liquid-disappearance-rate to increase or decrease. An experienced engineer will gain a feel for handling these changes, for optimum results. It may be noted that these experiments can be carried out with no great need for sophisticated sensors and monitors inside the borehole: the frequency and amplitude can be controlled from the surface, and really the only parameter that must be measured is the liquid-disappearance-rate, which can be measured at the surface.

[00210] It should be noted that, traditionally, when liquids are injected from a borehole into the ground, the liquid tends to spread outwards horizontally. Most ground material is highly stratified horizontally, and is much more permeable horizontally than vertically. Slosh-pulsing tends not only to reduce the fingering effect, by acting to homogenise the ground material between the fingers, but acts also to homogenise the ground material vertically. Thus, where traditional injection has produced an irregular horizontal disc of injected liquid, slosh-pulsing can be expected to produce an area of injected liquid that is larger in horizontal extent, very much larger in vertical extent, and considerably more homogeneous over the whole injected area.

[00211] In some installations, the borehole in which the apparatus is installed is remote from a producing well. The operation of injecting liquid, in that case, probably is being carried out in order to augment the pressure head of the liquid in the ground close to the producing well. In that case, liquid probably is being drawn off from the remote producing well during the time slosh-pulsing is taking place in the borehole. Thus, the condition might well never be reached in which the inlet check-valve 145 only opens right at the end of the ram piston's up-stroke. In that case, the designer might decide that slosh-pulsing should continue, more or less until the producing well is finished.

[00212] In FIG. 15, as in many installations, especially where the borehole is deep, the exhausted air in the exhaust conduit 142 is still at a pressure that is very much higher than atmospheric. It may be arranged that the air compressor 148, at the surface, receives this still-partly-pressurised exhaust air as its intake, as this can be an important economy compared with simply dumping the exhaust air to atmosphere.

[00213] The pneumatic actuation and exhaust valves are shown combined in a single unit 135, but the valves may be separate structures, if desired. The valves may be controlled and operated electrically, or by other means, as desired. It will usually be required that the operational conditions produced by the valves be monitored, and e.g pressure sensors, flow-rate sensors, displacement sensors, etc, may be provided, as required.

[00214] The designer will prefer to keep such items to a minimum, since the pipes and wires associated therewith all have to be led to the surface, and have to be found room for, within the tube 125. The ram 128 and associated components shown in FIG. 15 may be located right at the bottom of the borehole, and there may be no other rams over the rest of the depth of the borehole. In that case, probably, there is ample room for providing a number of monitors, etc. But in other cases, there may be several of the rams, at different depths within the same borehole—in which case each of the rams must be arranged to provide room for a route for the conduits, pipes, wires, etc for the rams below, which might be rather difficult in a borehole of around ten cm diameter.

[00215] It will be understood that, in FIG. 15, the actuation piston and the ram piston are combined in a single piston structure 130. The actuation cylinder 136 and ram cylinder 129 are likewise combined in a common physical structure. These pistons (and cylinders) may be separate, if desired, and connected by suitable structures for transmitting forces therebetween. For example, it might be arranged, especially if the borehole is of shallow depth, for instance, that the pneumatic actuation system components be at the surface, and a mechanical strut be used for transmitting forces down to the ram piston, which is located at the depth of the aperture. However, if the borehole is deep, the length (and weight) of such strut would mean that the operation of the ram would be sluggish and not fully controlled; preferably, the pneumatic actuation components should be located close to the ram components, whereby the pressurised air stored in the actuation conduit is immediately available, right by the ram, for precise operational timing and movement of the ram.

[00216] In FIG. 15, make-up liquid is supplied, from the surface, via the inlet conduit 143. Instead of the separate pipe, as shown, the conduit may alternatively be provided as the interior of the tube 125, i.e the whole tube may be full of the liquid to be injected. The inlet liquid may be pressurised, or may have the natural pressure of its hydraulic head, as required. If the ground material around the borehole, when remediation commences, is depleted of liquid, i.e is unsaturated (which it often will be, if the invention is being made use of), liquid may pass into the borehole at large volumetric flow rates, at low pressures. The higher pressures and smaller rates occur during the later stages, including the slosh-pulsing stage.

[00217] As mentioned in relation to FIG. 15, the piston 130 can be actuated via a mechanical strut, the pneumatic operator being positioned some way above the ram. FIG. 16 shows such an arrangement. A pneumatic operator (not shown) resides at the top of the strut 150, and drives the piston 152 downwards to initiate a cycle.

[00218] The inlet control valve 153 may be left open. The inlet check-valve 154 closes as the piston 152 descends, and liquid is forced out of the chamber 156, and out of the borehole. When the piston 152 is later raised, the check-valve 154 opens, and the chamber 156 fills with liquid from the inlet conduit 157. Later, when slosh-pulsing has commenced, as described the pressure in the chamber 156 remains higher than the inlet pressure, so the inlet check-valve 154 remains closed until the piston 152 has risen partway back to the start position.

[00219] In FIG. 16, the inlet control valve 153 may alternatively be closed, just as the piston 152 starts down. Thus, on the downstroke, the piston is forced downwards, not only against the resistance of the liquid in the chamber 156, but also against the vacuum that arises in the isolated space 158. The benefit of doing this arises when the piston reverses, that the inlet-check-valve 154 remains closed, as the piston rises, even if the pressure below the piston drops below the inlet pressure. The fact that the inlet-check-valve remains closed during the piston up-stroke means that the vacuum can assist in sucking liquid back into the borehole, and thus assist in starting, and continuing, the slosh-pulsing stage (in which the liquid is sucked back, and has an inwards velocity).

[00220] In FIG. 17, the ram piston 160 is actuated mechanically. A rod 162 extends to the surface, where it can be raised and lowered. The piston 160 includes an upwards extension 163, by means of which the piston can be pulled upwards, by the rod 162, against the upper stop 164. The tube 165 is full of the liquid to be injected into the ground. The liquid may be under pressure, and, as in FIGS. 15, 16, the liquid may simply flow out of the chamber 167, and into the ground, so long as the ground accepts the liquid. The piston 160 need not be actuated during this preliminary phase.

[00221] To actuate the piston, the rod 162 is dropped. The rod has a tapered bottom end 168, which engages the tapered socket 169 in the piston. The engagement forms a seal, whereby liquid from the chamber 167 below the piston cannot escape into the inlet conduit 170 above the piston. The weight of the rod resting on the piston increases the pressure within the chamber 167, whereby the liquid is driven out into the borehole.

[00222] After a volume of liquid has been discharged, the rod 162 is then raised (from the surface). The rod and piston rise together, until the pressure of the liquid in the chamber 167 has dropped below the pressure of the liquid in the inlet conduit 170. Then, the rod parts from the piston, and liquid flows into the chamber 167, from the inlet conduit 170. The rod pulls the piston up, until the components are once again at the position as shown in FIG. 17.

[00223] The fact that the rod strikes the piston a sharp blow sends a shock wave through the liquid in the chamber below the piston, and this shock wave can emanate out from the borehole, into the ground material. It can be highly useful, in some cases, to create a shock wave, and then to back that up immediately with a continuing increased pressure in the liquid. The effect of the shock wave can be to dilate the pore-spaces in the ground, at least momentarily, as it passes through the ground. The benefit arises if the steady increase in pressure (due to the weight of the rod resting on the piston) might result in some of that increased pressure being trapped in the dilated pores before the pores can settle back to their normal smaller size.

[00224] Sometimes, the rod will bounce when it strikes the piston, which can give rise to several shock waves. Dropping the rod does not always increase the rate at which liquid can be injected into the ground, but sometimes it does, in which case the effect is worth the extra wear and tear on the components that results from the dropping. The rod 162 may be some hundreds of meters long, whereby the weight of the rod, and its impact on the piston, are considerable. By contrast, the piston 130 in FIG. 15 suffers no impacts at all during operation, and this design should be favoured where operations lasting several months (and perhaps millions of cycles) are contemplated. It is also possible, in FIG. 15, to actuate the valve 135 in such manner as to profile the pressure changes, e.g a sudden high pressure at first, followed by a steady pressure—although it is not possible in FIG. 15 to equal the violent shock wave produced by dropping the rod in FIG. 17.

[00225] The liquid being injected might be oil, where the borehole is, or is being restored to be, a producing oil-well. Where the borehole is remote from a producing oil-well, the injected liquid might be water. The system as described might alternatively be used generally for injecting water into the ground, and especially for injecting water that carries dissolved or suspended treatment chemicals, for groundwater treatment and remediation purposes. In some case, the requirement might be to inject liquids other than oil or water, and the invention might be contemplated in those cases too.