Devices designed specifically for ultra-sensitive vibration sensing offer advantages for existing products such as those used in the oil and gas industry that employ a “gap change” model. The new HP micro-electromechanical sensor (MEMS) accelerometer architecture was designed with increased sensitivity and small form factors in mind, as well as very low drift characteristics and the robustness to operate in high- and low-energy environments. Initial designs are capable of withstanding more than 2,000 g forces and even greater transience. This architecture offers several usage possibilities.

By leveraging the very low drift characteristics of the MEMS, this sensing technology was initially mooted as an inertial navigation chip capable of tracking and identifying high-value items in global supply chains, geo-ring fencing, and other location-based applications where absolute positioning systems were unavailable and relative positioning based on origin were appropriate. One suggested application was to leverage the sensing technology to allow operators to “recreate” lost land pipeline maps. Capturing the coordinates of each pipe in three dimensions with very limited disruption to operations is a solution to a vexing problem. Advances in high-density solid-state memory means the storing of sensed data is not necessarily constrained by the need to transmit that data immediately.

MEMS, sensor, HP

MEMS inertial sensor employs variable capacitor transduction. (Images courtesy of HP Labs)

Wireless sensing systems also can be used to detect vibration changes in rotating and reciprocating equipment to help better identify changes in equipment behavior, a core tenet of condition-based monitoring. Engineers can retrofit equipment with wireless sensors at multiple points to show changes developing, not at discrete points, but across the entire equipment real estate as harmonics migrate and precess with time.

The same principle of introducing external webs or sensor nets capable of detecting small changes in induced behavior applies equally to static infrastructure such as valves, risers, and flowlines. By tuning the sensor to specific monitored frequencies, the possibility of deducing flow regimes within closed systems at any point in that system becomes of real value.

Analytics
Sensing is, of course, only part of the story. With increased sensing comes increased data volume. Challenges exist today in detecting and deriving value from current data volumes; multiplying that volume by orders of magnitude can be limiting. However, complementary research into the potential transmission demands or local storage/retrieval of generated data and the aggregation and interpretation of large numbers of simultaneous data streams offsets this limitation, which will result in easier and faster deployment.

HP, wellhead, production

Visual analytics displays wellhead flow regimes in three-phase productions.

Given the limitations in coverage of any process/equipment, the most effective approach would be to characterize the various operational modes the process exhibits, correlate this to observed or known conditions, work back through the data to look for indicators of onset of change, and then mine incoming data for those or similar patterns.

Understanding the implications of sample intervals and aliasing is important. The primary risk is that the patterns indicating change onset occur at such a radically different interval to the sample interval that, in effect, no indicators exist within the data.

In tests, three-phase flow through a wellhead monitored for pressure volume temperature exhibited variable flow “regimes” under common choke settings. The only variable was fluid composition, and though patterns and correlations were difficult to extract, the flow regimes became more obvious using visual analytical tools.

Using this ability to exploit sparse sensor data and extract visibility into the underlying process, then using firsthand knowledge and correlated data to devise the root cause, provides a means of establishing the character of the processes to be monitored.

Drilling domain
Incorporating an inertial sensor into the bit assembly during drilling may seem overly speculative given the conditions to which it will be exposed. The challenges of drilling extended-reach and maximum reservoir contact programs demand accurate geosteering, and while measurement-while-drilling techniques provide a strong guidance, precise and accurate bit location contributes significantly to successful programs and has incremental benefits in reducing non-productive time.

The challenge is to design an embedded sensor capable of inertial measurement/guidance at one extreme, while also providing data feeds that help indicate bit wear and the immediate geological conditions encountered. This needs to be to the point of indicating when and where a particular interval or bedding plane is encountered, or for inclined drilling, when the bottom edge of the bit encounters first contact with geological interval. This requires a sensor that can operate in harsh environments yet still offer very high sensitivity levels.

As discussed above, designing the MEMS for thermal stability, and using a surface capacitance approach (vs. closing the gap), results in a robust sensor that can withstand very high g-forces – e.g., more than 2,000-g resilience coupled with 100 nano-g/root Hertz noise floor – while measuring fine changes. The construction of single-crystal silicon extends the operating temperature range even higher, up to 392ºF (200ºC), which enables its use in high-pressure, high-temperature wells.

Although the physical limitations of the sensor’s stability and resilience may be overcome, the harsh and noisy environment means that the signal (direction changes) is extremely small compared to the noise (bit vibration and impact). Detecting subtle change in a noisy environment is not unique to the oil and gas industry, and can be resolved by applying techniques such as those described above to demonstrate how minor trend changes can be isolated from overwhelming background noise with great precision.

With a flat frequency response in the base design, long wavelength monitoring offers an intriguing possibility in the domain of both exploration and production. Though this scenario requires geotechnical modeling and verification, the possibility, in certain circumstances, to detect the periodic shifts in gravity caused by large semifluid bodies moving under the influence of tidal forces would suggest gross fluid volumes and imply permeability, even fluid composition and petrophysical conditions.

The challenge here is to implement a relatively dense array of low-cost, low-power gravimeters based on six-component accelerometers measuring lateral and vertical gravitational effects.

This concept may only be feasible in carbonate reservoir structures and would be constrained by many parameters such as hydrocarbon composition, water, gas, oil, and condensate ratios. However, applying a simple difference model to monitoring offers potential for this class of sensors, enabling the measurement of minute gravity field changes as the fluid composition changes.

Ultimate numbers, rates of proliferation, and cost benefits remain to be determined, but the inevitable result is that sensors are getting smarter, more capable, and available for deployment at more cost-effective price points. The potential for commercial benefit through sensor deployment and subsequent results interpretation is significant, and in the oil and gas industry, the challenges, needs, and drive exist to see continual breakthrough in the development of, and returns from, advanced sensing developments.