Remote monitoring and control works for unspectacular wells in ordinary fields.

Field automation has historically been biased toward keeping the upper crust of wells - the very highest producers - in top form. Systems for automated monitoring and control of such wells have set the bar high, specifying real-time data, full supervisory control and a growing investment in centralized resources to help transform data into knowledge and action.

While this data-intensive solution is optimal for top producing wells, it is economically unattractive on a field-wide basis. These resource-intensive SCADA systems ignore the more modest needs of the average well, where data that arrives at the right time, rather than in real time, provides just the heads-up that's needed for careful maintenance, timely intervention and more strategic field management.

Average wells

For average wells, the economically viable solution lies in the middle ground between SCADA systems and the wireless "smart dust" of the future. Smart dust, with its promise of low-cost remote monitoring via armies of tiny, autonomous sensing nodes, offers a stark contrast to the centralized structure and high costs associated with SCADA solutions. Yet even as smart dust researchers re-think costs and re-distribute autonomy within the monitoring network, the resulting nodes, if and when they are commercialized, will likely be very limited in the range of actions and reactions they can support.
Fortunately, what the average well needs, a technological middle ground, is here today in the form of smart-sensor technology. Like SCADA, these sensors support the relay of production data from every well to a central repository. Like smart dust, they push the limits of compact size and wireless, intelligent, self-powered configurations.

But smart sensors for the wellhead also break new ground in two important ways. First, although they monitor constantly, they parse the effective dataset transmitted to the central system, thus avoiding the data glut created by SCADA systems. While SCADA systems' remote nodes monitor and store data every second and send huge packets of data to the central server every 10 or 15 minutes, smart sensors can report a more manageable dataset on a once-a-day basis. Second, the sensors have enough native intelligence to make selected optimization decisions on their own, rather than acting as "dumb" devices that only monitor and relay the data.

By the end of each day, these intelligent sensors have a rich sample of pressure, flow rate, tank level and other measurements to transmit, documenting the day's production, alarms and any interventions taken. The average well gets just the amount of automation it needs, and engineers get just the right amount of data.
This combination of breakthroughs, made possible by a series of patented technologies applied to the typical well, is rewriting the economics of field automation and production optimization. To understand why, it helps to look both at the economics of the average well and the massive scale of the data created by the traditional automation project.

Right-time data

Of the 860,000 wells covering the onshore United States, only 5% produce more than 100 boe/d. The majority of those high performers are automated. Almost all the rest still rely on pumpers to fill out gauge sheets for their daily production data, which can take weeks to reach the engineer's desk.

The missed opportunity is tremendous. The benefits of near-real-time data - in this case, daily data from yesterday's production - have been well established. These benefits hold even for wells producing as little as 10 b/d, as proved by the handheld-based data capture devices currently in use on more than 12,000 wells. While these solutions fall short of real automation, they do illustrate the widely documented benefits of daily data: faster attention to the wells that need it, smarter coordination of daily priorities and, ultimately, production increases that are regularly in the range of 2% to 5%.

Why haven't operators clamored to adopt automation solutions that provide such benefits? The culprit is the high overhead cost associated with monitoring, storing, relaying, normalizing and reporting real-time data, which is the industry default due to the focus on the high-performing well. To support real-time data, the central server must have ever-growing capacity, software and other resources for intensive data management. Remote devices need computing power sufficient to capture tens of thousands of measurements in a typical day and communications capacity to relay this data to the central server. Such requirements commonly call for trenching at the well site to install the necessary electrical and communications cables. Both upfront and ongoing costs become prohibitive for the average well.

Intelligent monitoring via smart sensors drastically rewrites the cost equation by targeting right-time data from the onset. Right-time data, also known as relevant-time data, is by definition the data available in just the right amount, and at the most suitable time, to support the most meaningful action or decision possible. For most wells, for example, a daily production summary in a graphical report is the right-time
data to enable the next day's priority setting for the pumper. For a tank that has developed a leak, an alarm feature alerting a pumper or engineer immediately is right-time data.

Unlike SCADA systems, the new generation of smart sensing devices does not assume that real-time is always right-time. Smart sensors save costs by monitoring constantly, sampling and storing data frequently instead of continuously, and by delivering data daily, unless an exception condition warrants multiple relays in a single day. What represents "constant" monitoring varies by device, according to its purpose. For example, the intelligent electronic-flow gas monitor is designed to take a reading every second, while monitors for tank levels and pressure each record a measurement every minute. In each case, the time intervals for monitoring, sampling and relaying the data are defined by the analysis and actions that can be undertaken based on that particular data.

The result is a new generation of wellhead monitoring devices with everything needed to provide right-time data in the lowest cost, lowest energy configuration possible. All the devices are self-powered, wireless and completely integrated with a small solar panel, bi-directional radio/satellite transmitter, battery and a 100-MIP processor. As a case in point, the gas-flow monitor described above has a footprint of 4 in. by 4 in., costs 50% less than a typical real-time gas meter, and it requires less than an hour to install since there is no wiring required.

Gas field optimization

A mature gas field in Texas provides an excellent case study to demonstrate the potential impact of right-time data. The regional office of St. Mary Land & Exploration in Shreveport, La., operates wells in 20 different fields across Texas, Arkansas, and Louisiana, using 13 contract pumpers. During the past 4 years, they used right-time data in two key time scales, daily and monthly, and found it substantially impacted both production and operations in the two Texas fields where it was implemented.

St. Mary uses a handheld-based data capture system for a waterflood unit covering 22 wells in the Clarksville field, and for another 36 producing wells in the Box Church field. The first wave of benefits came from daily data, which immediately revealed allocation issues ultimately leading to more accurate well-level production reporting. The new system also exposed production issues, such as liquid loading, much earlier and thereby shortened response time to correct the problems.

For periodic field-level planning, right-time data was viewed on a longer time scale. With months of pressure, production and flow rates, and with a higher degree of confidence in the accuracy of well-level producing rates, St. Mary had the data in the format and density it wanted for reservoir simulation and other analytical techniques - a first for the Clarksville field under St. Mary's operation. Following the analysis, St. Mary was able to map out operational changes and complete several workovers. Production has outperformed forecasts consistently, resulting in increased production of 30,000 bbl in the past 4 years.

Although this kind of right-time data, gathered by pumpers and transmitted via handheld devices to a central server, improves operations measurably, it runs into limitations with time increments more granular than once-a-day summaries. Intelligent monitoring devices fill the remaining gap economically, supporting situations when real-time data is right-time data, both by handling alarms and by intelligently reacting to on-the-spot measurements with corrective logic. Smart sensors thus offer actionable, right-time data to inform, and even initiate, responses that optimize production.

Imagine, for example, a gas field with liquid loading problems but outfitted with smart sensors programmed to issue alerts when exception conditions are met. The appropriate sensor could even be programmed to help optimize plunger lift on its own.

In a traditional automation system, a remote device would identify and relay the plunger lift problem to the central server. There, an engineer would be alerted, and probably a pumper would be sent to "tweak" the unit. Several repeat visits might be necessary before the optimal setting was found.

With smart sensors, an alert could also be generated easily. But the devices could do more. They could actually be programmed to test, via trial-and-error within a limited range, different settings for plunger cycle times, recording the results and self-diagnosing when a certain setting is yielding optimal production rates.

The onboard processor in a smart sensor, and its built-in programmability, offers a myriad of possibilities for tailoring intelligent monitoring and responses to the known challenges of the field. Optimizing plunger lift is just one example. While traditional automation is built on sending all the data back to the server for the "smart" decisions to be made, intelligent monitoring re-distributes the ability to respond among both centralized and remote resources, introducing new possibilities for economic production optimization.

Conclusion

The miniaturization of well monitoring and automation componentry is well underway. It is set to revolutionize both return-on-investment (ROI) economics and data management requirements for the 95% of onshore wells currently without automation. Smart sensors are changing the economics of upfront and ongoing costs, meeting ROI hurdles even for small wells, and enabling lower-cost implementations of field-wide automation and optimization strategies.

Comments