HPHT Cementing GuidelinesHOW TO USE THIS DOCUMENT These Guidelines are an aid to assuring best practice in planning and execution of cementing jobs on HPHT wells. The aim is to provide the Drilling Engineer with a tool to prompt actions by others. This will usually be the Service Company Engineer, his laboratory technician (who has a substantial and key role), their back-up from other locations and specialists within their company, and the all-important Company Representative and the service engineers on the rig. The Guidelines are not intended to prescribe how the job must be done, but they should enable the Drilling Engineer to ask for evidence that a proper, technically assured process is in place and being followed. It is intended that these Guidelines are passed to the service company and that they ensure the issues raised are addressed and actions documented. The fact that some aspect or other is not addressed here should not be construed as meaning it is not important. Experience is very important in minimising risk. There are a number of key questions and these are indicated in italics. A listing of these Key Questions is also given in an attached Table. Areas covered are: • The Planning Process - what you need to be aware of from the start • Slurry Design - some guidance on what to ask and what to look for • Spacer Design • Integrated Job Design - mixing, pumping, mud properties, making sure it all comes together • On the rig - execution, what to look out for. It is often simple things which cause the major problems 1. Introduction HPHT cementing requires considerable additional vigilance on the part of all involved in the planning and execution because the consequences of a job failure are more severe than experienced in more normal wells. It has probably cost much more to get to where you are and, if it all goes wrong, it will cost a disproportionate amount to put it right. Investing time and effort in the detailed planning is vital to minimise risk. The following areas need careful attention: • Slurry design • Spacer design • Basic materials quality and suitability • Equipment selection • Liner Systems, hangers and running tools • Cement Mixing • Slurry Placement/mud displacement • Contingency planning: • in case of problems during the job • in case the job fails to secure the objectives, eg shoe leaks “Train-wrecks” in HPHT cementing operations are usually caused by simple things being overlooked or a complete, total, lack of knowledge about some aspect - (doesn’t know what he doesn’t know!). You cannot spend too much time looking at seemingly trivial details and asking simple questions about issues which seem minor. (As an example; it may seem impossible on such a critical job that cement will be moved from one silo to another without the rig Cementer being aware of the move, but it isn’t, and it has happened, and it will happen again and the consequence can be catastrophic). 2. 2.1 The Planning Process find out what experience the Service Company has locally with HPHT wells? Involve the Service Company early, 3 months prior to spud is reasonable: What experience do the individual engineers have with HPHT? 2.2 if the HPHT experience of the service company engineers is limited, establish how experience will be brought in. One way this might be done is to have an experienced HPHT service company engineer from another region come and review both the planning procedures and the cementing programme. Also, consider getting a second lab to verify the pilot slurry design(s), e.g. a lab of the same service company in a different region which has routine HPHT work. What support is going to be brought in? 2.3 the experience, commitment and competence of the lab staff will be crucial. Getting valid lab test data for HPHT is very time consuming and far in excess of that required for less critical wells. If the lab is weak, your exposure to risk is increased substantially. How can the lab demonstrate that it is more than adequate? 2.4 the cementing service company must interact with the mud company and with the equipment suppliers. Darts, Plugs, liner hangers, running tools, sealing surfaces, packers, elastomers, etc, all require extra attention. Make sure everything will work at the conditions you have and with the fluids you will be using. Don’t address these aspects in isolation from each other. What commitment is there to provide more than a slurry and a pump? 2.5 ensure roles and responsibilities are clear and are understood. There many different ways of running the operation, with more or with less responsibility devolved by the BP DE to the service companies and drilling contractor. Just make sure everyone knows what they are responsible for. What are the roles and responsibilities? 2.6 2.6.1 Is the rig good enough? the rig compressed air supply for bulk cement handling - check that the air driers are working properly. Is there a history of bulk supply problems? Moisture introduced into the bulk cement can seriously compromise the slurry design and lead to very confused lab results because of differences between rig samples and lab held samples which have not been through the same exposure. This is particularly true for blends, ie G+35%. How and when will these issues be addressed? 2.6.2 is the cement unit up to the job? Do you need to up-grade the HP fluid ends? What is its record of failure on ordinary jobs? What assurance can be given? 2.6.3 are critical slurries to be batch mixed? Can the volumes be accomodated? Do you need more than one batch tank? Can it be fitted on the rig? What are the limitations imposed by the rig, logistics and weather? 2.7 Temperature: this is the parameter which has the most influence on the slurry design and it is vital that the maximum amount of information about expected temperatures is laid out and made available to all involved at an early stage. Lab testing must be bracketed around the most likely temperature. The starting point for cementing is the undisturbed geothermal temperature - not log temperature, not MWD temperature. These temperatures obviously have a role to play in the estimation of undisturbed geothermal temperature but do not themselves form the basis for estimating cementing temperatures. Both software modelling and API Cementing Temperatures (BHCT) need to be viewed with caution. A coherent picture of the most likely cementing temperature environment needs to be built up from as many inputs as possible. What are the most likely temperatures, how are (were) they obtained, what confidence can be placed on them, what can be done while drilling to enhance confidence and reduce uncertainty? Who owns the issue and understands it? 2.8 Hydraulics: Deatiled, thorough planning is absolutely critcal for wells with a narrow pore-fracture window. Rheologies of fluids will require special attention including special lab measurement at HPHT conditions. It is frequently the case that computer simulations of the job create considerable, indigestable quantities of results and do not clarify the available options but obscure them. To avoid this the cementing service company engineer should summarise outputs onto a spreadsheet to indicate trends. An example would be to produce a single XL Chart plot of maximum downhole pressure against displacement flow rate for three different mud rheologies. This is easily understood as a chart with three lines but will normally be presented as 30 sheets of A4 print-out. Is there a clear understanding of the risks? Surge pressures and pressures to break circulation must be addressed - there must be a good integrated approach agreed by the mud company and the cementing company. How will the mud company and the cementing company work together to understand what can be done to optimise the cementing process? 2.9 Contingency Planning: Again, much more attention is required than for non HPHT wells. Lab testing to obtain a suitable KO plug (say) will take several days. Make sure that contingency slurries are developed as the well progresses. Do not wait until unscheduled events arise. For example, setting 9 5/8” casing deeper can have a profound effect on the slurry design and may take a week of lab testing if pilot studies are not already underway. Have plug designs ready for shoe squeezes and Kick-offs Have contingencies been identified, is there a documented and agreed plan of what is needed, by whom and when? 2.10 Assurance Process: Is everyone clear what the process is and how it relates to them? Are the interfaces satisfactory? 3. Slurry Design The key areas to be considered in designing an HPHT cement slurry are listed below and then each aspect is addressed: • Bulk cement sampling • Temperature • Cement selection • Additive selection: • Retarder • Fluid Loss Aid • Weighting Agent • Silica • Other additives • Slurry Sensitivity The main aim should be to “keep it simple”. Avoid secondary additives or additives to counteract the action of others, eg anti-settling aids, retarder aids. 3.1. Bulk Cement Sampling: HT wells will require silica, most commonly dry blended silica flour. The lab will work with samples from three sources: • samples of cement and silica which they weigh out and mix themselves • samples taken from the bulk blending plant • samples brought back from the rig For various reasons, these can all give different results when tested under the same conditions. To minimise risk it is vital that correct sampling procedures are rigorously followed. Again, this is a point which cannot be overstressed. Imagine a busy lab, under pressure to get results. It has only two rigs but many samples. The samples are all in the correct type of tin can. The lids are all marked with the rig name, the date, the silo number, etc. Under pressure to get this mixed and that mixed, this report done, that file found. The lids get switched…. You think it dosen’t happen? It does. Containers must be moisture proof and airtight, full and uncontaminated by anything. It sounds basic but many times compliance is less than satifactory. The consequence can be thousands of feet of 9 5/8” casing to drill out. Details should be on the container side - not on the lid. No job should proceed without a test on rig samples which is consistent with tests on other lab samples. 3.2. Additive Sampling and QC: For HT, the level of control on additives should be increased. Early in the lab design stage specific batches of additives should be “earmarked” for the project and set aside in sufficient quantity that there is no need to mix batches (manufacturing runs) of additives. This minimise variability due to additive QC. Rigorous tracking of Batch Numbers needs to be in place and specific products at the rig site should all have identical Batch Numbers 3.3. Temperature In high temperature wells, the standard API temperature equation for determining cement circulating temperature (BHCT - the lab testing temperature) from the bottom hole static temperature (BHST) should be viewed with caution. The data set on which this equation is based is very limited for HPHT wells and the slurry design is usually sensitive to small changes in temperature. Also, the API equation is not appropriate when the thermal gradient exceeds 1.9oF/100 ft (35oC /km). Therefore it is recommended that for HT wells (BHST>300 oF, (149 oC)) and/or the thermal gradient >1.9oF/100 ft (35oC /km) “Welltemp”, or an equivalent heat transfer model, is used to predict cementing test temperatures. In exploration wells where thermal gradients are less certain, computer modelling should also be used to interpolate logged temperatures back to BHST. The more data and modelling, the better the most likely temperatures can be bracketed and the appropriate lab designs chosen. 3.4. Cement Selection: Not only do different cement manufacturers’ products behave differently at HT, cement from the same manufacturer, manufactured at different times, will exhibit different behaviour. This is due to the complex chemical nature of the product and inherent variability in the raw materials being used. For normal well cementing, an Well Cement which meets API Specification 10 should be acceptable. However, the API Specification does not address suitability for HT conditions. When the temperature exceeds 300 oF (149 oC) BHST and/or slurry density exceeds 2.1 SG (17.5 ppg), the choice of bulk cement will have an impact on slurry performance and the risks associated with slurry design. The reactivity of some cements can make them very difficult to retard at high temperature, and the use of so-called “difficult to disperse” cements will impact slurry rheology and amount of weighting agent required to create a pumpable slurry. Wherever possible in HPHT cementing, a recognised high quality well cement with a track- record at such conditions should be used. This aspect needs to be addressed early in the planning process. Changes at a later stage can be problematic. Cements with non-linear response to retarder concentration should be avoided. 3.5. Additive Selection: The following is specific to HPHT productive intervals. 3.5.1. Retarders In general, normal lignosulphonate retarders are not appropriate (e.g. D13/D800/D801/ R12L/R-10L/ HR-4L/HR-6L) for wells where the BHCT exceeds 230 oF (110 oC) and the use of highly refined HR-12/HR-12L/HR-15/R15L/R-15/R-23L/D28) or synthetic or chemical retarders e.g. D109/D110/ R-14L/SR-30/ SCR100/HR-25 is preferred. 3.5.2. Fluid Loss Control Fluid Loss Control is necssary when the slurry is used in a narrow annular clearance (<1.5”) where a permeable productive formation is exposed. Where the annular clearance is below 1” then fluid loss control is necessary to prevent bridging whether a permeable zone is exposed or not. For high temperature application, the latex based fluid loss additives are widely used by Schlumberger-Dowell; the advantage being that latex does not viscosify adversely. However, latex HPHT slurries can be complex with up to 9 additives in a slurry. This complexity can bring its own problems and is best avoided. (Example latex based additives are D134/Latex 2000/BA-86L). Other high temperature fluid loss aids (D143//FL-32L/Halad 100//Halad 413) tend to increase the viscosity of the base slurry. The advantage with these designs is often simplicity and cost, however the disadvantage is increased ECD and some mixability issues. The newer additives (FL-45N/Halad-600LE) do not viscosify as much but have similar costs to the latex systems. 3.5.3. Weighting Agents The use of cement weighting agents should not be required below 2.05 SG (17.1 ppg). The use of dispersants to achieve a satisfactory “reduced-water slurry” is recommended. Effectively there are three possible weighting agents where higher densities are required: barite (4.3 SG) (D31/Barite) haematite (5.0 SG) (D76/Hi-Dense/Hematite) manganese tetroxide (4.7 SG) (D157/Micromax/W-10) Barite is the most readily available and lowest cost but is not suitable due to particle size and usual chemical impurities. Haematite is the most common weighting agent. Where slurry volumes permit the use of batch mixing, the haematite should be added to the batch mixer and not pre- blended with cement. Where pre-blending has been used, separation of haematite during pneumatic transfer has been observed and obtaining a representative sample of a 3 component dry blend presents a further difficulty. The density difference and coarseness of the haematite requires careful slurry design to prevent haematite differentially settling from the slurry. Manganese tetroxide appears to be the most suitable weighting agent and has the advantage that it can be added to the water before the cement, its fineness prevents rapid settlement simplifying the mixing operation. There has been recent concern about the use of the material because of variability in the particle size and impact on slurry sensitivity (particularly retarder response). There has also been concern about resistance to acids, particularly its ability to act as an oxidising agent. Currently it is not believed these issues should prevent use of the material. The issue of variable particle sizes should be managed by particle size analysis on materials supplied. A further system in use is Dowell’s product “DenseCrete”. This is a blend of cement/haematite and other dry bulks to give a carefully designed range of particle sizes. These blends appear to be more stable during transport and have better rheologies then the equivalent cement haematite blends. 3.5.4. Silica Where the BHST exceeds 230 oF (110oC) cements exposed over periods (months - years) will exhibit the phenomenon of strength retrogression. This is caused by a solid state phase transformation at high temperature due to the calcium - silica ratio producing a weaker, more permeable, material. To prevent this, silica is added to the slurry design to produce a more thermally stable product. The permeability deterioration is particularly relevant to HPHT Gas Wells. It is also important for producer wells to consider flowing temperatures during peak production. Strings far above and close to surface may require silica Two forms of silica are normally available: silica flour (D66/SSA-1/S-8) silica sand (D30/SSA-2/S-8C) The minimum concentration of silica added to cement should be 35% however studies have shown advantages when using silica sand of raising this to 50%. When using silica flour, the material must be preblended with the cement. This requirement adds a significant risk for the following reasons: • when the initial blending is done, the bulk is often not fully homogeneous and sampling errors can occur • with subsequent transfers, the blend becomes more homogeneous, but may also undergo some changes (aeration) due to moisture in the compressed air transfer system Therefore,samples taken from a blend at the yard may not match those taken at the rig. A 5% variation in the flour content of a blend can have a substantial impact on the retarder response and requirement. Where large slurry volumes are to be pumped, preblended silica is only practical option. Where slurry volume permits batch mixing, the use of sand in tote bins is preferred to minimise the risk of sampling and ensuring and accurate silica content in the final slurry pumped. 3.5.5. Other Additives Other additives used in HPHT slurry designs include: • Dispersants (for mixability and rheology) (D80/D65/CFR-3L/ CFR-3/CD31/CD-31L) • Anti-foam (to reduce air entrainment during mixing) (D144/NF-5/FP-14L) • Anti-settlement aids(to prevent solid and/or liquid separation (Silicalite97L/FDP-C-533/D153/LW8-L) • Synergistic additives (which may support other additives e.g. retarder aids) (D121/D94/Component R) 3.6. Slurry Sensitivity HPHT slurries can be very complex, and in most cases additives are being used at the upper end of their design limits. As a consequence the slurry properties are much more likely to be influenced by changes in: • temperature • mixing/shear • additive concentrations • density • additive order of addition • contamination The impact of each of these on the slurry design should be assessed. 3.6.1. Temperature: Sensitivity of the slurry to Once the lab design temperature has been determined, the thickening time should be obtained for 15oF (8.3oC) hotter than BHCT. Should the thickening time reduce below job time the thickening time of the proposed slurry should be increased. The compressive strength should also be determined for a test temperature 15oF (8.3oC) lower, or at liner top BHST whichever is cooler. If no set is seen in 36 hours retarder selection should be re-addressed. 3.6.2. Mixing/Shear Particularly in the case of batch mixing, cement slurries can be affected by both the time they are held on surface after mixing and the mixing energy they receive. The standard laboratory mixing procedure for cement slurry testing is defined in API Specification 10. When a slurry is designed for an HPHT application, it should be mixed and then held at room temperature and pressure simulating the holding time expected on surface. Some cements have been shown to have 6 hour pump times at high temperature but to exhibit significant viscosity rise when held at low temperature and pressure. Any rise in consistency above 30Bc during this surface conditioning should be sufficient to eliminate that slurry design. In an extended mixing period, the slurry may be continuously circulated through a recirculating mixer where it will see intermittent period of high shear. For extended mixing times, where large batches are being mixed, the slurry may see more mixing energy than during laboratory design. The intermittent nature of the shear has been demonstrated to have more impact than when all the mixing energy is applied in a single continuous mixing interval. The final slurry design should be tested either with a simulation of the shear the slurry will experience, by estimating the energy input during batch mixing, or using the following simple mixing programme: Mix for 1 minute at 1,000 rpm Mix for 20 seconds 12,000 rpm Mix 1 minute 1,000 rpm Mix 20 seconds 1,000 rpm Mix 1 minute 1,000 rpm Mix 20 seconds 12,000 rpm Mix 1 minute 1,000 rpm Mix 20 seconds 12,000 rpm This is equivalent to about twice the mixing energy received in a standard test, some slurries can reduce thickening time by >30% using this procedure; behaviour which needs to be avoided. 3.6.3. Additive Concentration The thickening time should be determined for retarder concentrations +/- 5% of that in the expected design. Acceptable slurries have: pump time > job with 5% less retarder pump time < 2 x base slurry with 5% more retarder pump times which decrease with less retarder and increase with more retarder. In the field, when using a batch mixer, the additives should be gauged from calibrated pails weighed on an appropriate scale. If the slurry is to be mixed continuously, it is advisable to use a cement batch tank to prepare the mix water, unless assurance can be obtained on cleaning the pit and lines to prepare mix water. In most cases the number of additives excludes use of an LAS, but if it were to be used a calibration check should be completed on volumes indicated before use. 3.6.4. Density It is critical that the slurry is pumped at the designed density for both well control and slurry performance. Pumping the slurry light may increase the susceptibly to settlement of the weighting agent, pumping it heavy may reduce the thickening time. The slurry should be pumped after checking with a pressurised mud balance (calibrated at the appropriate range). 3.6.5. Additive Order of Addition Slurry sensitivity has been shown to be affected by different orders of addition of additives. To prevent this affecting performance, the laboratory should document the order used during slurry design and this should be included on the slurry recommendation, and followed when preparing the mix water in the field. 3.6.6. Contamination The impact of contamination by mud and/or spacer can have a substantial impact on slurry pumping time and rheology. The thickening time of the slurry should be confirmed with a 10% contamination of both mud and spacer. The high solids content of high density spacers can reduce pump time of the slurry even at low levels of contamination. If this occurs in the lab, it is recommended that a similar concentration of retarder is included in the spacer. Should the mud have a dramatic effect on pumping time, spacer should be pumped ahead of and behind the plug to prevent mixing. Additionally, care must be exercised when pulling out from the top of liner to minimise fluid mixing. 4. Spacer Design On HPHT wells, both water and emulsion spacers have been used (the later favoured by Dowell with OBM). The main areas of concern with spacer design are: • spacer stability • selection of weighting agent • rheology • compatibility 4.1. Spacer Stability The ability of the spacer to support the weighting agent on surface and under downhole conditions should be confirmed. This is particularly critical where pore and fracture gradient margins result in spacer density being very close to mud weight. The ability to support the weighting agent at surface should be confirmed in the laboratory by leaving a volume of spacer static for 2 hours in a measuring cylinder using the Free Water procedure used for testing Well Cements (API Specification 10). Prior to pumping, the density in the pit should be confirmed with a pressurised mud balance. The stability downhole should be assessed using the BP settlement tube and determining any free fluid after leaving for a minimum of 4 hours under downhole conditions. 4.2. Selection of Weighting Agent The options of weighting agent are: • Barite • Manganese Tetroxide • Haematite • Brine 4.2.1. Barite Barite is the most common material used and is recommended where the required spacer rheology can be achieved. However, the stability of the spacer must be checked. 4.2.2. Manganese Tetroxide Will produce the most stable spacers with the best rheology but at more expense. Manganese Tetroxide has been used in conjunction with Barite to produce better spacer rheology and stability. The fineness of the material results in spacers that are almost stable without any polymer viscosifier. 4.2.3. Haematite Not recommended unless spacer density precludes use of other weighting agents formulating a stable spacer is more difficult. 4.2.4. Brines as With the advent of the formate brines it has become possible to formulate low solids or solids free high weight spacers. The advantage of turbulent flow at low flow rates and good stability would make these an option where fracture gradient and pore pressures are close. Whilst these have not been pumped in HP/HT wells to date they should be considered for future applications. Compatibility testing should be thorough and rigorous. 4.3. Spacer Rheology In many cases the ability to achieve turbulent flow of the spacer, or a density hierarchy between the spacer and the drilling fluid is compromised, due to the small difference between pore and fracture pressures. Where one of the above cannot be achieved, rheology of the spacer and pipe movement will be the only available methods for achieving good mud displacement. To optimise spacer rheology it should be determined at simulated downhole temperatures (along with the mud) to ensure effective mud displacement. Surface rheology measurement should also be taken to provide a quality control step on the rig. 4.4. Compatibility It is critical that the compatibility of the spacer is confirmed with both the mud and the cement. Using a range of mixtures (as a minimum 72/25-50/50-25/75), the rheology should be determined at BHCT (or 180oF if BHCT not practical). The also 10 minute gels should also be included. If the 100 rpm is 25% greater than the highest 100 rpm reading of the uncontaminated fluids, the spacer design should be reviewed. If plastic viscosity/yield point or 10 minute gels suggest the slurry may become unpumpable, or separate into one or more phases, the spacer design should be changed. In addition, the Thickening Time of the cement with spacer contamination as described in above must be determined. 5. On the Rig: The attention and vigilence of all those on the rig is vital. Again, small, seemingly trivial mistakes can adversely impact the job. Experienced personnel who know what to look for and can work around unexpected difficulties are vital if risk is to be minimised. Some particular aspects to look for are: 5.1. before delivery of any bulk cement blends, ensure that the receiving silo(s) is emptied and thoroughly cleaned. The system should be thoroughly checked, filters, vents, water traps, etc cleaned. Someone should specify just what should be done. This should be agreed by the Drilling Contractor, the Service Company Engineer and the Cement Engineer on the rig. Someone should sign-off just what has been done. there must be a rigorous, documented system maintained by the Barge Captain, or other responsible person, which unambiguously records movements to, from and between tanks physically check all Batch Numbers on all materials to be used. This activity should be signed-off Mud Engineer(s) and Cementer(s) together check and record Chloride content of Drill Water prior to addition of additives. Is the Cl- content consistent with what the lab have used? Has there been any new delivery of Drill Water? 5.2. 5.3. 5.4. 5.5. a “brain-storming” session should be held on the rig prior to running casing/liner to capture all possible ways the job may be jeopardised during execution. The points raised should addressed and responsibilities assigned if there is a risk of having insufficient personnel to cover all aspects during job execution, bring in extra assistance, eg bring the lab guy who has worked on the slurry written contingency actions once cement pumping has started should be in place and should have been discussed and agreed if batch mixing is to be used, the lab testing will have been done to take account of this. The lab simulated time for batch mixing should be stipulated to the rig and if this time is exceeded, slurry should be dumped. mix water should not be prepared before absolutely necessary. If mix water preparation starts but a delay occurs beyond a few hours, dump the mix water. software simulated pressures/rates should be compared with the actual during the job - not just afterwards samples of blend, mixwater, additives, drill water, spacer(s), mud must be taken, fully labelled and held securely until the job is accepted as fully satisfactory for displacement volumes, remember the tolerances on id, uncertainties on pump efficiency (if using rig pumps) and mud compressibility minimise the time to pull above the liner once the plug has bumped but do not swab the well. Keep pipe in motion, pulling, and monitoring weight at all times. It is better to keep moving slowly if at all possible unless weight is increasing 5.6. 5.7. 5.8. 5.9. 5.10. 5.11. 5.12. 5.13.