WaringWorld

Gas Lift Austria

2012-03-05T12:24:00.001-05:00

Just returned from a two-week event in Austria, my third such event.

This one was special given that it was about gas lift wells and that we had to do a review of 150 wells. We focussed our efforts on 90% of the production which comes from the continuous lift wells. I wrote a macro to create PROSPER files from an completed file and a spreadsheet of data. We performed a standard analysis of each well, looking for opportunities to increase production by a combination of changing lift gas rate, lifting deeper or changing the tubing size. It was very important to include the 1-2 km flowlines in this analysis. Even though the wells are working at a high water cut, we managed to find opportunities to increase oil production over 20%.

During this event we toured the gas lift equipment repair shop. We saw an example of creating conventional (not side pocket) mandrels from pup-joints.

Shop built conventional gas lift mandrel

Also, we reviewed parts from their retrieved gas lift valves and found amazingly little wear. These valves had been operating for several years and the stems and seats were almost like new. The field has sales quality (dehydrated gas) and quite low gas lift injection rates, but still that was unexpected. They did not have any special unloading procedure and only use working (IPO) valves, no orifice valves. The valves stay in the ground for 5-10 years, typically.

Automated IPM GAP Building

2012-02-04T13:23:00.000-05:00

For the past 3 weeks I have been working with a client in Houston on a new onshore field. The field will eventually be more than 1000 wells. Fields this large with this much equipment can be modelled in Petroleum Experts IPM Suite, in GAP. However, the process of building such models can be tedious, especially since the wells are very similar and so are the facilities. The client had already put together a macro to build the PROSPER models. I spent much of the past three weeks creating a macro to build the entire GAP model. The first step is to create templates for similar equipment configurations, then build portions of the GAP model from a list of well and facilities names using these templates.

The middle layer of a three level hierarchy of a large GAP model.

I also arranged the model in flowsheets with a three layer hierarchy to make navigation and viewiing much easier. One good thing about creating a model purely from data and a macro is that with a little work different views of the model could be created, for example an x,y coordinate layout or a hierarchical tree layout. Probably the best thing is that all the elements of the model are created consistently, with proper naming and the correct values. Small errors do sometimes creep into models that are built manually.

Such a large project requires as much automation as possible. The client already has a great plan for building an digital oilfield solution, including automated surveillance support systems. Petex IFM and IVM feature in this plan. The first of many IFM workflows are alredy built for PVT matching and FBHP calculations.

We also put together the beginning of a training plan and support model. Multi-discipline work will be very important to the success of this project. Already we have had several interesting conversations between Reservoir Engineers, Production Engineers and Facilities Engineers. It is always great when you can see the team starting to really work with the same data, tools and understanding.

2011-12-25T13:43:00.000-05:00

Merry Christmas to All!!

PSO Sustainability

2011-10-28T15:18:00.001-04:00

I was in the Middle East again this week, doing a post investment review of a major software installation that was put in place to help the Production System Optimization process. The software was installed with great fanfare in 2006 and lasted about two years before it rapidly fell out of use. We interviewed 15 people and none of them (NONE!) said the software was bad or didn't work. So, what happened?

Someone recently told me that the secret to longevity of technical work is re-selling. Perhaps you designed the process or purchased the software. You sold the software and processes to your customer/bosses and they bought it. You have to keep selling them on it, again and again. On the face of it, that's crazy. But, reality is that people change. Staff change. Managers change. Organizational charts change. And, folks simply get bored. So, what is logically a one time decision is really an un-ending series of challenges to that decision, and you the provider of that process or software have to keep showing up to the debate, keep bringing your salesmanship A-game.

I wish this company or software installation were an isolated incident, but its not. I can think of 3 other products that had the same thing happen to them. If you want your PSO processes and software tools to last in your organisation, here are some suggestions.

Establish Permanent, Full-Time Technical Leader (Focal Point) for Managing the New Process
Employ an energetic, senior technical person who has experience and can lead large projects. Ensure succession with a similarly qualified individual. Put them in-charge also of reviews of related hardware (e.g.metering and instrumentation) and given the resources for this.

Agree and Mandate a Standard Operating Procedure for Surveillance Process
It is very important to create a process in the company, by the company, that can work with the software the company has access to, and make it detailed enough to be able to train staff and establish KPIs for it.

Conservatively Estimate the Staff Time Required, and Staff Up
Think your process or software will make things more efficient? Make sure that everyone is actually doing the work already, which they may not be, which they may not admit. Going from not doing work to doing work takes more staff time.

Use a system of Super-Users in the organization to work directly with the staff involved in the process, with effective succession planning (every 3 years). Maybe it can be done with new graduates following a well-defined process. Following a well-defined, mandatory process would make it possible to make that process more efficient, sparing time for improvements elsewhere.

Set and Check KPIs
Establish KPIs and expected values, routinely monitor and report KPIs and respond to non-compliance. These are all standard ingredients in sustaining a process.

Use the Full Range of Change Management Techniques over the Long Term
This includes management understanding the procedure and routinely being interested and asking questions of staff, re-selling of the process and tools to staff and management, dealing publicly with non-compliance.

Ramp-Up Slowly, Learn from First Implementations
Do not follow the normal process of piloting and then broad implementation. There is no need for piloting, the benefits are proven. Start with one asset, learn from it, adjust, and move to the next asset. Maintain progress. Expand staff as progress is made.

Nothing Above would be a ‘Project’ with an End Date.
PSO is a process. Do not make this a project. This is the start of your new process of enhanced PSO that uses the best tools by competent staff and with ongoing management support. It only ends when the field no longer produces.

Re-Sell the Product to the Same Folks
Return periodically with the KPIs and the estimated value (either an actual measurement or at least multiple anecdotes about how well the process or software is working). Make your pitch again. What support can you give them? What suggestions to the users have? Praise the work done and give constructive criticism along with the support.

If you have some further ideas, leave some comments!

Tunisia

2011-09-30T11:24:00.003-04:00

OK, enough summer. Now back to work as global oil and gas consultant. This time I am in Tunisia, which is a wonderful country.

Burney in the Sahara

I led a Produce-the-Limit type production system optimization review lasting two weeks with about 25 staff.

It took a day each way to get to the field close to the Libyan border in South Tunisia, which is in a barren, rocky, sandy desert. Luckily it was hot, but not horribly hot.

Desert view from Twin Otter plane

The people I worked with were very nice, worked hard and are very interested in additional production. Tunis is a beautiful town, or at least there are a great many beautiful views to be found.

Concorde Hotel Restaurant view

Together we found some additional oil production through the expanded use of gas lift and the addition of ESPs. We also were able to quantify the uncertainty of their production rates and created a forecast and manpower requirements for additional wells they will drill. The team was very pleased with a detailed plan to increase production in the short term by 20%-30% and in the next few years by perhaps 100 or 200%.

Pipeline running into the sunset

As I am packing to leave, I am hoping to come to Tunisia again someday.

Summer Sailing

2011-07-11T15:49:00.000-04:00

I'm taking a couple months off this summer to enjoy the family. I hope you can do the same.

Cheers!

Landing on a beach, Jordan Lake

Wyatt takes a break from the helm.

Petex User Meeting

2011-06-16T13:46:00.000-04:00

I am attending the Petroleum Experts User Meeting in Edinburgh. Edinburgh is such a great city, very historic but comfortable with friendly people and many great sights.

This is an annual event put on by Petroleum Experts, the dominant provider of software for well and integrated system modeling for the oil and gas industry. I have used Petex software since the early 1990s and for several years taught PROSPER, the well modeling software. My last 5 years with Shell was full time promoting and managing this software suite globally for Shell.

It was great to see lots of former colleagues and friends.

Left to right: Burney, Mark Deere, Abraham Ekebafe, Hamid Guedroudj, Leigh Fry, Harry Sommerdijk, Thankgod Egbe, Umoh Magdalene Mussah, Khatan Ansari, Mike Pincock, Fidelis Tendo

If you think you know about the IPM suite and the state of the art of integrated modeling in the industry, you might want to keep watching what Petroleum Experts are doing. There have been many improvements lately and they just keep getting better. This year they have included in the next release major improvements in how they model CO2, salts, beam-pump wells, and coal-bed methane wells. Several of the packages are getting major improvements in the interface. Soon you will be able to make a forecast and for every point in the forecast see a pressure gradient. RESOLVE, which orchestrates several packages within the suite and also third-party software into a common model, has always had scripting required to do interesting things. RESOLVE now has visual workflow creation, which looks much less intimidating than a screen full of programming code. Even more radical is that RESOLVE will be able to build models made up of small parts of the IPM applications and will be able to use objects the user creates out of these small parts in multiple models. This is the equivalent of re-usable code modules but made out of applications. So, a system can be made which calculates a forecast where at every time step a minimum fluid level in each well is calculated and pump speed adjusted. Or, a system can be made that looks uses a mixed-integer optimizer to find the best of multiple development option combinations based on full-field lifecycle economics. Taken to its logical conclusion, the system could be used to create any oilfield software product that has as its core an overlap with what IPM does.

Finally, Petroleum Experts has bought NSI, a company that specializes in visualizing real-time data. Their previous product, OFC, is now called IVM and will link to IFM to help quickly create visualizations of the massive amounts of data that come from the oilfield after that data has been filtered and manipulated through engineering software.

The capability of the larger suite is already being demonstrated, and the potential of linking and reusing modules both in project development and in brownfield operating environments is huge and growing.

Austrian Development Optimisation

2011-06-12T06:05:00.000-04:00

This month I went back in Austria to lead another event. The weather was pretty nice when I arrived and there were a lot of pleasure craft on the Alte Donau, an isolated section of the Donau (Danube) river in Vienna.

I led a team of 16 production engineers, reservoir engineers, geologists, operations engineers, process engineers, project managers and others focused on discovering opportunities in the development of a relatively new field. The field already has some producing wells and has about the same number of wells already planned. The team's goal was to identify a subsequent set of new wells, improvements to the existing wells and the facilities modifications that would be required. We found 25 subsurface opportunities that could be pursued in 12 options, combinations of surface and subsurface scenarios. Also, unlike a normal PtL, economics had to be run on each opportunity before the scenarios could be assembled into those scenarios and ranked. I put together a macro to automatically run all the economics once a table of data had been assembled.

Shell originally came up with the idea of a 'creaming curve' (a derivative of Pareto cost-benefit analysis) to show the benefit additional exploration wells have in terms of cummulative discovered oil. The curve normally shows less and less benefit for more and more exploration wells. We use this in Produce-the-Limit type events to show management what they would lose in benefit if they cut a portion of the capital costs spent on a particular project, and which portion should be the first to be cut.

In this case I turned the preferred case into a creaming curve of NPV vs CAPEX and included the overall and incremental NPV/CAPEX as separate curves. You can see from this that the last opportunities on which you would spend CAPEX are to the right in the curve, these have the lowest incremental NPV/CAPEX (gold line) and lower the overall NPV/CAPEX for the project (red line). The result is a flattening of the total NPV vs CAPEX (blue line) as the CAPEX increases.

The currently planned wells will double the current production. In this event we found sufficient additional economic opportunities that the company will be able to again double the planned production level, still at a low upfront investment level.

Kazakhstan Focussed Optimization Event

2011-04-22T01:42:00.003-04:00

I've never been to Kazakhstan before. Aktau sits on the Caspian Sea, but it is not much of a resort town. The area around it is desert and the city was once the site of a nuclear reactor used to create plutonium for the USSR. It now is one of the hubs for oilfield work in Kazakhstan, and there was chatter that the place has radioactive dust.

Luckily, I was there when the weather was pleasant. The temperatures can range from -30 to 40C (-22 to 104F).

Kazakhstan is the 9th largest country in the world, and it is very, very empty.

The field we reviewed produces a waxy crude of about 50API gravity. That is some nice crude! The wells flowed naturally but the field desparately needs water injection for pressure support and due to the low productivity, needs ESPs for artifical lift. The small plant runs well but staff numbers and operating costs are very high. It was an unusual use of an optimization event to check assumptions and plans already agreed. Largely thanks to Paul, in addition, we provided opportunities to save OPEX and CAPEX and uncovered some safety and integrity gaps, for example pig launchers without proper safety interlocks and materials not rated for high pressure CO2 service. We identified 20% gains in production immediately at no cost, but additional gains up to 40% can only come later, once the major projects are in place.

Paul and I got to ride in a Mi-8-T.

Mi-8-T

Wikipedia says the Mi-8-T is the world's most produced helicopter, but I can say that the one we flew in vibrated like no helicopter I've been in. It was such a level of vibration that my skin would itch in whatever areas would reach harmonic vibration with the machine. I had to brace myself carefully, not touching certain parts of the seat in order to stay sane for the hour flight. The helicopter did save us from a 14 hour, round trip ordeal on gravel roads to the field and back, so it was fully worth the temporary irritations.

Paul 'relaxing' during the flight.

Romanian Focussed Optimization Event

2011-04-22T01:31:00.001-04:00

Romania is a different kind of place. There is rapid development everywhere, but also the kind of place that your great-great-great grandparents might recognize. Sheep and shepards, manual working of the earth, horse carts, simple living. All alongside high speed internet, urban/industrial sprawl, traffic jams and satellite dishes.

Farmer, horse, wellhead

Romania is the first place to produce and refine oil (http://www.150deanidepetrol.ro/history.html), starting in the 1840's. Today production has fallen off quite a bit. The oil industry is actively modernizing after years spent in a Warsaw Pact economy.

Along with my old ex-Shell colleague, Paul Wood, we reviewed a medium-sized gas field which had a fascinating mix of oil and new technology. Russian compressors installed in 1977 (based on 1960's technology) pumped gas while consuming hundreds of liters of lubrication oil per day were just down the road from a modern centrifugal compressor station. The separation station has no automated level controls on its 10 separators--they all used manually operated water dump valves! But for dehydration, they used a molecular sieve.

Brainstorming opportunities on the asset's cartoon

For the first time in my career I needed continuous translation in order to work together with the local team, which slowed down the effort a bit. But we still managed to review all the wells and find some gains. The larger gains require modernizing the surface facilities. The team came up with a clever re-use of one of the modern compressors (currently used as a standby unit) as a low pressure suction compressor. The project is expected to give about 15% extra production and 10% increase in recoverable reserves, with a nice profit. The event showed how multi-discipline teamwork, an open mind and a focussed process can uncover substantial value.

We stayed at Hotel Europeca in Craiova. The food was pretty good, but you might be put off by the menu:

I liked Romania. The weather was pleasant. The infrastructure was good for a developing country and the people were pleasant and eager to help. I'm looking forward to returning.

Austria Focused Optimization Event

2011-02-25T08:44:00.002-05:00

I've been in Austria now for two weeks. Beautiful country (I imagine it is nicer when the sun is shining). Nice people.

Center of Vienna

Oilfields? Yes, there are, and they have quite a few wells. The field I worked on has been on production since the 1950s and is very well run and well organized. The equipment has been maintained in excellent condition.

A new sucker rod pumping unit

Land of wheat, wine and wells

I led a team of 12 engineers (Reservoir, Production, Process, Geological, Operations, and Maintenance) on a quest to find opportunities to increase value in the short/medium term. I used a process similar to a Produce-the-Limit-type review, if you know what that is. The event lasted 2 weeks and identified an 80% gain in oil production rates that will last for several years. Even though it means upgrading most of the field's production, injection and electrical systems, the project is still profitable.

It was in many ways a textbook example of optimization. The reservoir in this case has a lot of potential. The wells were limited by the chosen artificial lift technique and the surface systems were already at their maximum. One discipline working alone, or a small team working on this as a normal course of their job would not have been able to create such a quantum step forward. What is required for success in such an event is an agreed process, support from the leadership, challengers external to the asset, time spent preparing the team to have an open mind and accept challenge, 100% dedication to this event for two weeks by all the team members, a field visit as a reality check, a complete understanding of all of the system constraint levels, and someone who has 'been there before' to lead them through all of this (me).

In the end we performed a full well review, estimated all the upgrade costs, produced a 35 page summary and technical document, and (of course) a Powerpoint pack of 50 slides, including a notional 2.5 year, 6-phase project plan, and project economics. Plus we had a look at embedding PSO in their daily business.

The company has agreed that this was a great success and many more such events are being planned.

Gas Lift Courses Completed

2011-02-02T07:50:00.004-05:00

Larry and I finished teaching three courses of one week each. The material ranged from basic to quite advanced and the skills exercises are mostly based in WinGLUE. WinGLUE started as a tool I wrote while in Shell to do gas lift surveillance, analysis, design and optimization. With the further development and commercialization of WinGLUE by AppSmiths, it has evolved to do those things and many other things that have to do with gas lift including BHP survey analysis, gas lift valve reliability, links to automatic valve testing equipment, incorporating valve performance from the Valve Performance Clearinghouse and lately, WellTracer.

In general the students did quite well. It is always gratifying to watch them learn, especially watching the 'Aha!!' moments when they finally understand a difficult concept.

One of the more challenging topics was the analysis of WellTracer (CO2 injection) surveys, which are very useful for finding out lift depts of gas lift wells. WellTracer finds leaky valves and tubing leaks above the annular fluid level. It has even found casing leaks in a few wells. WellTracer is in use by several companies and there have been over 300 of them run in the last few years. Gas lift is very flexible and often gets very inefficient rather than failing outright, causing lost production and wasted injection gas. WellTracer is an inexpensive technique that causes zero deferment and using it requires zero intervention and and so zero risk of dropping tools in the well as is the case with wireline. As such it can be used on the full range of wells including your best wells (where no one will let you risk wireline work) and the worst wells (where instability in the well makes wireline work impossible). AppSmiths includes the analysis of WellTracer results and recommendations for the well in the cost of the survey, in most cases.

At the end of the classes Larry and I got pretty good evaluations, including:

"The class was awesome!"
"Burney and Larry are very knowledgeable instructors. They are real experts."
"I learned a lot in the class and not only about gas lift."

Teaching is hard work. That sort of feedback makes all the effort worthwhile!!

Dinner of Technical Gurus

2011-02-01T07:38:00.001-05:00

In Oman we had a nice dinner out at the Mumtaz restaurant. It was great food, but also a meeting of great minds. It seemed that we were even greater later, after a few beers.

Larry, Jan-Willem, Burney, Clement, Chandran, Jeff

Back in Oman

2011-01-28T09:53:00.003-05:00

January finds me again in Oman, one of my favorite places, and teaching, one of my favorite jobs.

Burney teaching gas lift

This time I'm sharing the load with Larry Peacock of AppSmiths Ventures.

Larry Peacock of AppSmiths Ventures, LLC

The weather was rainy when I first arrived. Very unusual for rain to continue for several days in this country. Finally the sun came out and Oman turned into the paradise that it normally is this time of year.

The coast of Muscat, Oman

Larry and I are teaching three classes on the basics of gas lift and how to use WinGLUE. The participants are a mixture of Production Technologists (aka Production Engineers) and Production Programmers (similar to office-based Operations staff).

Gas lift class in Oman

Although it has been about 9 years since I taught the course, it helps that I wrote the original course and was the technical advisor on the software development. Larry reminded me of many things and I taught him some more secrets of gas lift. Oh, and we are teaching 40 Omani's and expat staff how to do surveillance, analysis, design and optmization on gas lift wells.

This weekend I wrote a draft of a new gas lift test for the course. Here are a few of the topics covered on the course and the test:

Gas Lift Competency

Why gas lift is able to increase or sustain production
What factors increase production into the wellbore through the completion zone
How pressures in the tubing are influenced changes in produced fluid characteristics and mechanical changes to the well
How the two primary methods of gas lift work and when to select each
The three main gas lift objectives for gas lift, in order of importance
The benefits and problems of gas lift
The primary operational hazard that causes gas lift valves to fail prematurely and how to avoid this
The relationships of pressure, volume and temperature on gasses and how to apply these relationships in gas lift analysis and design
How the density of certain fluids change and how to estimate pressure at depth for various fluids
How gas is most commonly measured and the most common reasons for poor measurement
How to calculate the amount of gas that can go through a valve, what factors affect this rate, and how to size valves for a given gas throughput
How to estimate the lift depth of a well from a flowing survey, from surface pressures, from valve mechanics and/or from valve throughput,
The effect of well surging on welltest and pressure survey measurements
The four main causes of well surging and how to recognize and eliminate each
The main types of gas lift valves, how each of these types work, how they are installed and retrieved,
The primary mechanisms by which gas lift valves fail over time
The two main kinds of unloading gas lift valves and when to apply each
How to select, set and size gas lift valves
How to determine if a gas lift valve is open or closed using valve mechanics
Two mechanisms for calculating the throughput of gas lift valves and the pros and cons of each
The mechanisms by which gas lift wells unload and the primary causes of a gas lift string to fail to unload properly
How to determine to which depth a string of gas lift valves can unload
How to space mandrels and the four criteria for a good mandrel spacing
For a given injection rate, how to quickly determine the tubing pressure at any depth and the associated production rate. How to use this technique in practice for gas lift design
What is the surveillance process, why it is especially important for gas lift and tips on performing gas lift surveillance.

WinGLUE Competency

How to enter a new well into WinGLUE with all the necessary well data in only a few minutes
How to determine the optimum amount of gas lift gas to a well and the operating envelope to avoid the four causes of gas lift well instability.
How to most efficiently review all the data entered into WinGLUE for a well
Which models to use and how to calibrate a gas-lifted well tubing pressure model in WinGLUE using three main criteria
How various calibration parameters affect the well tubing pressure model
How to calibrate the well inflow model
How to perform gas lift well analysis in order to identify problems, estimate gains from improvements
How to incorporate real-time surface data into a dynamic WinGLUE analysis
How to redesign gas lift strings for current performance with an existing mandrel spacing for maximum performance and stability
How to create and adjust a new mandrel spacing for the life of a completion
How WellTracer works and how to use it improve surveillance and analysis in simple and very complex situations. How to get WellTracer results even with unreliable data.
How to use measured valve performance data to improve analysis in WinGLUE
How to work efficiently with IPM GAP and get the benefits of full network modeling using WinGLUE
How the Automatic Valve Tester can check gas lift valve quality and improve reliability of gas lift strings

Holidays

2010-12-25T19:28:00.000-05:00

WaringWorld, Inc. took a break in operations for the Christmas holiday. It was a great time, especially with the addition of granddaughter, Caroline.

Nomographs

2010-12-15T10:14:00.000-05:00

Have you ever made a nomograph? Probably if you are under 50, then the answer is 'No'. I had not made one, although before the personal computer age I did use them quite a bit. In the early days of computing mainframes were used to produce nomographs to allow engineers to easily produce designs without hand calculations or resorting to writing programs themselves for the mainframe. (Punchcards anyone? Actually it is ironic and fitting that my spell checker doesn't have the word, 'punchcard'!)

Thornhill-Craver Gas Lift Choke Sizing Nomograph

I think I should make it clear that I didn't design the nomograph. That was the task of someone long ago, but it did remind me how arcane that knowledge is. Just to reverse-engineer the nomograph given the equation was a challenge. At least I was able to update it with metric units and make clean copies for the gas lift class materials. The old scans were not easy to read.

Focused Optimization Event

2010-12-05T10:39:00.001-05:00

The past few days I have been designing what I'm tentatively calling a Focused Optimization Event. This "barrel-chasing" workshop is similar to what Shell calls their "Producing the Limit" event. I think BP developed something similar called "Production at the Limit" or somesuch (if anyone knows, please tell me).

Building a common understanding of the production system

The Produce the Limit events started out at 2 full weeks (including a field visit on the middle weekend). They were a great way not only to identify hundreds of opportunities to improve production, but also to build a common knowledge of the field and to improve teamwork. We all had a great time during these events. They were very hard work, but I think what made them fun was that we were doing real engineering 100% of the time and intensely working together--things that we didn't get to do on normal workdays.

Hundreds of opportunities to increase production

As with many things in life, the bean counters got involved and started trimming these 'expensive' events, first down to 1 week where they were still successful with enough preparation. Then they were shrunk down to 2-3 days and were fairly useless and not much fun. Of course, then they were dumped entirely. Somewhere along the way we realized that many of the opportunities found in these events were not implemented and the effort spent on realizing these gains was not easily sustainable in the organizational structure that was in place. Happily, this laid the foundation for a whole new initiative on continuous improvement which is much better.

I have a client who has not done any events like this and so the challenge will be to hold an event that is as successful as the original and also kick-starts the changes necessary to sustain the production system optimization process.

Operational Philosophy

2010-12-02T10:17:00.000-05:00

A question came up today regarding the establishment of "operational philosophy" for new fields including rules for welltesting frequency (in addition to government requirements), well and reservoir monitoring, reservoir management, smart controls and automation, etc. These are huge decisions for a field development team and it can be difficult to know where to start.

Some suggestions:

a) By far the most valuable thing in the field are the reserves of oil and gas in the ground. Look at their reservoir recovery processes. Ask if something were going wrong in the recovery of those reserves and you did not know it, what specific information would help you understand the problem and what would you have to control to correct the problem? Repeat this for all the common things that can go wrong with your kind of reservoir given what you are certain about. Then look hard at your reservoir uncertainties for other things that could go wrong. There are many, many kinds of problems, but here are a few examples: If there were a thief zone taking most of the water injection and short-cutting across to the producer, how would you know this and how would you stop it? If part of the reservoir has no water pressure support, how would you know this and how would you stop it? Having the right information and controls--that is the key not only to running the field to recover the most reserves but to optimize the value of the field while it is producing. Please don't make the mistake of designing your instrumentation and controls for only the 'base' case.

b) Regarding instrumentation and controls, go for redundancy. If you want to see something sad, find a complex field with broken meters and gauges and chokes and talk to the team that is trying to optimize it. They cannot trust anything and struggle to make the simplest of decisions. Gauges and meters fail routinely. What would you do if an instrument failed? Could you use other information as a back-up? Could you use other information to check a critical measurement? For example, a subsea gas lift meter might fail (very often, apparently). Can you use upstream and downstream pressures over a choke, or over a pipe section to give an indication of flow rate? Could you use a second downhole pressure gauge to backup the first one if it failed?

c) A steady stream of the right data does cost time and money. Even the best managers have to defend capital expenditures at all levels, so it is no good just to say "We need this extra meter." Justifying the costs is straightforward using value of information concepts. To start, play out a scenario in the field reservoir simulation of things going wrong with no correction (because it is undetected) and the same simulation with things under control and look at he difference times the probability that this scenario might happen as the expected value of having this information. For example, a waterflood with high voidage in one corner (because no one can measure SBHP there) vs the right voidage with extra downhole gauges. Or a field with a poor (random) distribution of lift gas (because no one can meter the gas properly) vs the right one with spool sections and valves for easy installation of replacement/redundant meters. These expected values will likely be much larger than the costs of measurements. No one ever complained about a field where the data was too reliable and the controls worked all the time.

d) Make sure the gauges and meters that you get are themselves the most robust. When you have a choice, go for meters with wide ranges of application, that can be field calibrated, can be easily replaced. Sacrifice absolute accuracy to get robustness. Often the person who is buying the meter will ask questions about the likely flowrates, gas to oil ratios and water cuts. Don't give them single numbers. Give them the full ranges of values that are possible. For example, a single gas meter on a test separator sized for initial rates will be useless when the rates decline, or if the GOR is much lower than expected, or when the water cut climbs. It might be very undersized when gas lift is used on the well. And, give special consideration to gas lift measurements as they so often are poor (which is sad because dry gas should be the easiest to measure). Gas lift gas to each well deserves accurate measurement because it will directly impact the optimum use of the lift gas, but also because the associated gas from each well will be very distorted with poor gas lift measurements, which result directly in poor operational decision-making and therefore lost recovery.

e) Make sure that the data that you measure can be collected and managed. Data that is not readily available is effectively lost. Look at the data and apply a 60 second rule: can I get this data on my desktop and start to use it in less than 60 seconds. The longer it takes, the less likely staff is to make the effort to find out the truth and the more likely they are to make assumptions instead.

Newfoundland

2010-11-27T17:00:00.002-05:00

I had a visit to Newfoundland this week. St. John's is a nice town, with a port, right on the North Atlantic.

Port of St. John's

I got a walk from the hotel every day, past some very nice restaurants--surprisingly nice considering that St. John's is literally at the end of the earth.

Water street, St. John's, NFL, CA

If you get to go there try Portobello's (http://www.portobellosrestaurant.ca/) which was truly excellent.

Here is my quintessential Canadian photo...

Hockey after work!

...walking to work in the morning, carrying the hockey mask and skates. At the airport some girls were trying to get through the revolving door with hockey sticks laid horizontally across their cart. Hockey, eh?

My client had a variety of areas that they wanted me to review, including well testing, production allocation, production system optimization, and integrated production system modelling. They have a nice set of tools there including FlowManager (FMC) software, sold as a "Virtual Flow Metering System". The system is necessary for this client because the wells have to be tested at much different back pressures than the system to which they normally produce.

This was my first exposure to Flow Manager. If you are not familiar with this tool, it is a physical model running attached to a real-time system. The idea is to run the model with a known rate from a good welltest and calibrate the model using typical flow correlation values like friction and gravity coefficients, then have the program estimate the rates in real-time using the pressures and temperatures in the system. FlowManager did appear to take quite a bit of work to keep calibrated, and the client is working hard to get the Flowmanager models calibrated with a new set of high-quality data.

The client had a Petroleum Experts, Ltd. IPM model that had fallen into disuse. I selected the PE5 model from Petex which worked OK without matching, and re-ran all the lift curves with a better spacing of liquid rates. After that, we were able to get the GAP model calibrated again to a good set of relatively recent data and start using the model for optimizing in a number of different well routing and gas lift distribution scenarios. The IPM model will be useful for weekly production planning and daily production optimization.

I was able to leave them a report with instructions for simple calibration of the model with a new set of data, using the model to optimize against constraints, and archiving and sharing the model. I also gave them suggestion for improving their allocation and welltesting processes, as well as mentoring, and general teamwork.

PSO Event Planning

2010-11-17T21:17:00.001-05:00

I started making recommendations for a Production System Optimization event to be held in Europe early next year.

The challenge is to find a few hundred thousand barrels of oil.

Years ago, I participated in many of these events and they were always interesting and challenging. The object was to uncover opportunities to add value by relatively inexpensive changes to the production system. The key to success was to have a multi-discipline team, ringfenced for the event. We had these events that lasted from one to two weeks. The longer versions included a field visit where the team could verify and refine the opportunities on location, occasionally testing out ideas on-site. The event should be fun and I might even get to work with some colleagues I haven't seen in a few years.

Oman

2010-11-02T17:56:00.002-04:00

This week finds me in Oman, one of my favorite countries.

From the beach at Ras Al Hamra

It's not only that the scenery is gorgeous, but that the people are very friendly. In 40 years Oman has come from a medieval feudal kingdom to a modern society, thoughtfully led by His Majesty Sultan Qaboos bin Said. Oman is a well-run country with all the modern conveniences, but still retaining their essential cultural elements.

I was in Oman to teach a course, "The Business of Production System Optimization", that I developed while at Shell. The course teaches the process of PSO using a number of tools including a full field simulator. In the past I had used a full field simulator that I developed, but I also wanted to teach integrated production system modelling, so I redeveloped the course using Petroleum Experts, IPM software. In the course, teams compete for the most profitable set of optimization opportunities that they can negotiate for implementation. Their opportunities (combinations of pipeline cleanouts, workovers, re-routing, pipeline resizing, compressor re-wheeling, etc.) are run through IPM to determine the winner. In this course I had 17 students from 8 countries. We all had a great time and learned quite a lot about each other's operations. We ended the course with action planning to carry forward a number of valuable changes in the participants' companies.

Production and Injection Networks

2010-10-21T18:23:00.001-04:00

This week I spent some time working on an algorithm for solving a network of gas lifted production wells with the gas lift injection network that supplies the gas. One of my clients in the Gulf of Mexico wants to derive a system solution with spans both the production and injection networks in their sub-sea systems. The are using WinGLUE and are happy with the analysis they get on an individual well basis.

WinGLUE from AppSmiths, Inc.

The latest version of IPM does handle allow such a solution, but makes simplifying assumptions which would be appropriate for a large network. However, for a small network IPM does not take into account the gas lift valve performance. It looks like there is a simple mechanism to improve the solution that can be achieved with a combination of WinGLUE and GAP.

Construction Begins

2010-08-04T22:23:00.000-04:00

We broke ground on a major project in the backyard this week.

First step, remove the trees and grass...

The plan is to put in a pool and a garage. The construction crew says we might get to swim in October. I'm thinking that means Christmas.

Caroline

2010-07-24T22:11:00.000-04:00

First grandchild!

Adorable, a·dor·a·ble adj.

Delightful, lovable, and charming
Worthy of adoration.
Caroline.

Retirement

2010-05-05T21:57:00.002-04:00

Twenty-nine years went by fast. I had some great times, but all good things must come to an end.

My friends got together and gave me a nice farewell.

About 30 people came to see me off in New Orleans

They were nice enough to let me have the last word.

I give one more lecture

Leigh and Dawn

My great friends got me a cake.

My Team: Semi, Burney, Harry, Fidelis, Mike (my successor), and Aimara

And, I said my good-byes to my team, who traveled the world with me.

I miss them very much...

...but not all the time.