Controlling the growth of future LEO debris populations with active debris removal

Acta Astronautica 66 (2010) 648 -- 653

Contents lists available at ScienceDirect

Acta Astronautica journal homepage: w w w . e l s e v i e r . c o m / l o c a t e / a c t a a s t r o

Controlling the growth of future LEO debris populations with active debris removal J.-C. Lioua, ∗ , N.L. Johnsona , N.M. Hillb a b

NASA Johnson Space Center, Mail Code KX, 2101 NASA Parkway, Houston, TX 77058, USA ESCG/MEI, Mail Code JE104, 2224 Bay Area Blvd., Houston, TX 77058, USA

A R T I C L E

I N F O

Article history: Received 3 February 2009 Received in revised form 8 August 2009 Accepted 11 August 2009 Available online 2 September 2009 Keywords: Orbital debris Population growth Active debris removal

A B S T R A C T

Active debris removal (ADR) was suggested as a potential means to remediate the low Earth orbit (LEO) debris environment as early as the 1980s. The reasons ADR has not become practical are due to its technical difficulties and the high cost associated with the approach. However, as the LEO debris populations continue to increase, ADR may be the only option to preserve the near-Earth environment for future generations. An initial study was completed in 2007 to demonstrate that a simple ADR target selection criterion could be developed to reduce the future debris population growth. The present paper summarizes a comprehensive study based on more realistic simulation scenarios, including fragments generated from the 2007 Fengyun-1C event, mitigation measures, and other target selection options. The simulations were based on the NASA long-term orbital debris projection model, LEGEND. A scenario where, at the end of mission lifetimes, spacecraft and upper stages were moved to 25-year decay orbits, was adopted as the baseline environment for comparison. Different annual removal rates and different ADR target selection criteria were tested, and the resulting 200-year future environment projections were compared with the baseline scenario. Results of this parametric study indicate that (1) an effective removal strategy can be developed using a selection criterion based on the mass and collision probability of each object, and (2) the LEO environment can be stabilized in the next 200 years with an ADR removal rate of five objects per year. Published by Elsevier Ltd.

1. Introduction Fifty years after the launch of Sputnik 1, satellites have become an integral part of human society. Unfortunately, the ongoing space activities leave behind an undesirable byproduct: orbital debris. As of 1 June 2008, more than 17,000 objects were tracked by the US Space Surveillance Network (SSN). The majority of them, approximately 12,000, have their orbital elements maintained in the US Satellite Catalog. The cataloged objects are approximately 10 cm and

∗ Corresponding author. E-mail address: [email protected] (J.-C. Liou). 0094-5765/$ - see front matter Published by Elsevier Ltd. doi:10.1016/j.actaastro.2009.08.005

larger. Other than the 800 or so active payloads, this population is dominated by breakup fragments, spent upper stages, and retired payloads. The growth of the orbital debris populations has been a concern to the international space community for decades. Many policies and procedures are established to address the issue. A good example is the adoption of the space debris mitigation guidelines by the United Nations in 2007 [1–2]. However, recent numerical studies have shown that the debris environment in low Earth orbit (LEO, defined as the region up to 2000 km altitude) has reached a point where the debris populations will continue to increase even if all future launches are suspended [3–4]. The driver for the increase is mutual collisions among orbiting objects, a phenomenon

J.-C. Liou et al. / Acta Astronautica 66 (2010) 648 -- 653

predicted by Kessler and Cour-Palais [5]. In reality, the population increase will be worse than the “no future launches” prediction because satellites will continue to be launched and major breakup events, such as the Fengyun-1C (FY-1C) breakup [6] and Briz-M explosion [7], may continue to occur. Even with a full implementation of the commonly-adopted mitigation measures, the LEO population growth appears to be inevitable. To better preserve the near-Earth environment for future space generations, additional remediation measures must be considered. The concept of active debris removal (ADR) is not new, although there are major difficulties in the removal technique and the high cost associated with the actual implementation. Other issues, such as ownership, policy, and liability, also prevented ADR from being seriously considered in the past. However, the recent assessments of the LEO debris environment warrant a reconsideration of the option. From a modeling perspective, it is a straightforward task to examine the effect of ADR, and that is precisely the objective of the present study. The goals are to use the most recent NASA orbital debris evolutionary model to (1) develop simple, reliable, and objective ADR object selection criteria, (2) quantify the effectiveness of different ADR scenarios, (3) explore ADR strategies needed to stabilize or even reduce the future debris environment, and (4) provide guidance for the development of removal technology. The ADR modeling study was initiated by the NASA Orbital Debris Program Office in late 2006. The first effort was focused on the object selection criteria. A non-mitigation (sometimes referred to as the business-as-usual) scenario was used as the baseline for comparison. The main conclusion of the study was that the product of the mass and collision probability of each object was an excellent removal selection criterion (see also Section 2.2). Numerical simulations based on this criterion showed most objects in the critical inclination and altitude regimes were identified and removed, and the LEO debris population growth, using the non-mitigation scenario as a benchmark, was significantly reduced. These results were presented at the 2007 International Astronautical Congress [8]. The present study differs from the previous one in the following areas: (1) the tracked FY-1C fragments were added to the initial environment for future projection, (2) a more realistic scenario, where the commonly-adopted mitigation measures were implemented for future launches, was selected as the benchmark, and (3) the focus was on what would be needed to stabilize (i.e., no growth beyond the current levels) the future LEO debris environment.

649

ment based on a comprehensive NASA Orbital Debris Program Office internal database. Known historical breakup events are reproduced and fragments are created with the NASA Standard Breakup Model, which describes the size, area-to-mass, and velocity distributions of the breakup fragments [11]. The only exception to this process is the FY-1C breakup in January 2007. Since fragments from this event are very different from those of a typical breakup, their distributions are derived from the SSN tracked data [12]. The simulations described in this paper were completed in February 2008. Based on the catalog data available at that time, a total of 1536 FY-1C fragments had both good orbital elements and area-to-mass ratios derived from their orbital element histories and were estimated to be larger than 10 cm in size. Only these objects were included in the simulations. Although more catalog data became available later and additional 10 cm and larger FY-1C fragments were identified, the difference should not affect the overall outcome of the present study in any significant manner.

2.1. Benchmark scenario The future projection component of LEGEND covers 200 years from the end of the historical simulation. Future launch traffic was simulated by repeating the 1999-to-2006 launch cycle. The following postmission disposal (PMD) mitigation measures were implemented. Rocket bodies, after launch, were moved to 25-year decay orbits or to LEO storage orbits (above 2,000 km altitude), depending on which option required the lowest change in velocity for the maneuvers. In most cases, the 25-year decay orbit was the preferred choice for vehicles passing through LEO. The mission lifetimes of future payloads were set to 8 years. At the end of the mission lifetime, each payload was moved to either the 25-year decay orbit or to an LEO storage orbit. The PMD success rate was set to 90%. No explosions or deliberate breakups were allowed for future rocket bodies and payloads. Collision probabilities among objects were estimated with a fast, pair-wise comparison algorithm in the projection component. Only objects 10 cm and larger were considered for potential collisions. This size threshold is historically the detection limit of the SSN sensors, and more than 95% of the debris population mass is in objects 10 cm and larger. A total of 100 Monte Carlo (MC) simulations were carried out for future projection, and the averages were calculated for comparison.

2.2. ADR scenarios 2. Modeling tool The LEO-to-GEO Environment Debris (LEGEND) model is capable of simulating the historical and future debris populations in the near-Earth environment [9,10]. For this study, the historical component in LEGEND covers the period from 1957 to 2007. The model adopts a deterministic approach to mimic the known historical populations. To accomplish this, launched rocket bodies, spacecraft, and mission-related debris (rings, bolts, etc.) are added to the simulated environ-

The first step for ADR simulations was the development of target selection criteria. The following objects were not considered for removal since they did not significantly contribute to the growth of future LEO debris populations: objects smaller than 10 cm in size, objects with perigee altitudes above 2000 km, and objects with eccentricities greater than 0.5. In addition, operating payloads (assuming a nominal lifetime of 8 years) and breakup fragments were excluded from removal consideration.

650

J.-C. Liou et al. / Acta Astronautica 66 (2010) 648 -- 653

For any object i which was eligible for removal consideration, a simple criterion, Ri , was adopted [8]: Ri (t) = Pi (t) × mi ,

(1)

where mi was the object's mass, and Pi (t) was the collision probability of object i at time t based on a fast pairwise comparison algorithm [10]. The rationale of adopting Eq. (1) was straightforward. To maximize the effectiveness of ADR, objects with the greatest potential of contributing to future collision activities and generating the largest amount of debris must be removed first: these were objects with the highest mass and collision probability products. For ADR scenarios, the basic assumptions were identical to those in the benchmark scenario. In addition, active debris removal was implemented at the beginning of each projection year starting from the year 2020. Once the Ri values for objects with non-zero collision probability were calculated, they were sorted in descending order. A pre-defined number of objects with the highest Ri values were removed from the simulated environment immediately. Two scenarios, with removal rates of two and five objects per year, respectively, were completed (these two scenarios are referred to as PMD+ADR02 and PMD+ADR05 throughout the rest of the paper). Again, each scenario included 100 MC runs. 3. Simulation results Fig. 1 is an updated version of the results of a “no future launches” scenario where the historical environment was extended to 2007, including FY-1C fragments based on the catalog data. The sharp increase in 2007 was caused by the FY-1C breakup. Many of the FY-1C fragments have high areato-mass ratios, and at least some of them are likely to be multi-layer insulation and solar panel pieces [12]. The curve

beyond 2007 represents the average of 100 LEGEND Monte Carlo simulations. The decrease in population during the 20 or so years after the breakup is primarily due to the decay of these short-lived fragments. Overall, the trend of the curve in Fig. 1 is identical to that obtained from previous studies [3]. Even without future launches, the LEO population growth appears to be inevitable. The main result of the ADR simulations is summarized in Fig. 2. The top curve shows the benchmark case. Even with an implementation of postmission disposal measures at a 90% success rate and 100% explosion suppression, the LEO debris population, for objects 10 cm and larger, will increase by about 75% in 200 years. However, with a removal rate of just two objects per year, starting in the year 2020, the population growth can be reduced by half (middle curve). With a removal rate of five objects per year, the future LEO debris population is kept approximately constant for the next 200 years (bottom curve). Each curve in Fig. 2 is the mean of 100 MC runs, providing a good way to examine the “average” behaviors of different projection scenarios. On the other hand, each one of the 100 MC runs represents an equally probable outcome of the future projection. The spread of the 100 different MC outcomes also provides insight into the projected population growth. The distributions of the LEO satellite populations (objects 10 cm and larger) at the end of the 200-year projection from all 100 MC runs, of two scenarios, are shown in Fig. 3. The LEO satellite population at the beginning of the projection is 10,719, as indicated by the vertical line. It is clear that without ADR, the LEO population is expected to increase by an average of 75%. With a combination of PMD and ADR05, the mean population in 2206 is close to the current one, with a possible spread from 6000 to no worse than 18,000 objects. The spread of the PMD+ADR02 prediction is between the two shown in the figure.

Effective Number of Objects (>10 cm)

LEO Environment Projection (average of 100 LEGEND MC runs) 14000

Total Collision fragments Explosion fragments

12000

Intacts + mission related debris

10000 8000 6000 4000 2000 0 1950 1970 1990 2010 2030 2050 2070 2090 2110 2130 2150 2170 2190 2210 Year

Fig. 1. The simulated LEO debris population. The historical component is based on recorded launches and on-orbit fragmentations, including the FY-1C breakup. The future projection is the average of 100 LEGEND Monte Carlo simulations based on the “no future launches” assumption. At the end of the 200-year projection, the mean of the total is 12,595, with a 1 −  standard deviation of 3213 and an error in the mean of 321.

J.-C. Liou et al. / Acta Astronautica 66 (2010) 648 -- 653

651

LEO Environment Projection (averages of 100 LEGEND MC runs)

Effective Number of Objects (>10 cm)

20000 18000 16000

PMD PMD + ADR02 PMD + ADR05

14000 12000 10000 8000 6000 4000 2000 0 1950 1970 1990 2010 2030 2050 2070 2090 2110 2130 2150 2170 2190 2210 Year

Fig. 2. Comparison of three different scenarios. From top to bottom: postmission disposal (PMD) only, PMD and ADR of two objects per year, and PMD and ADR of five objects per year, respectively.

35 PMD

Number of MC Occurrences

30

PMD + ADR05 25 20 15 10 5 0 0

5000 10000 15000 20000 25000 30000 35000 10 cm and Larger LEO Populations in 2206 (100 MC predictions)

Fig. 3. LEO satellite populations (objects 10 cm and larger) at the end of 2206 from 100 different MC runs. The current population is indicated by the bold vertical line. The 1 −  standard deviations of the PMD and PMD+ADR05 distributions are 4287 and 2443, respectively.

The averaged cumulative numbers of collisions predicted by the three scenarios are shown in Fig. 4. A dashed, straight reference line is added for comparison. The PMD result shows an obvious deviation from the straight line. For the PMD scenario, a total of 45 collisions are predicted for the next 200 years. The predicted numbers of collisions are 35 and 27, respectively, for the PMD+ADR02 and PMD+ADR05 scenarios. The positive deviation from the linear increase for the PMD scenario is an indication of collision instability in the LEO environment.

To quantify the effectiveness of each ADR scenario, a simple parameter, effective reduction factor (ERF), can be used [8]. It is defined as: ERF = [total no. of objects reduced in 2206]/ [no. of objects removed via ADR through 2206]

(2)

The ERFs for the two scenarios are summarized in Table 1. For the PMD+ADR02 scenario, the parameter is about 10. It means that for every object that is removed (via ADR) from the environment, the total debris population at

652

J.-C. Liou et al. / Acta Astronautica 66 (2010) 648 -- 653

LEO Environment Projection (averages of 100 LEGEND MC runs) 50 PMD

Cumulative Number of Collisions

45

PMD + ADR02

40

PMD + ADR05 35 30 25 20 straight reference line

15 10 5 0 2000

2020

2040

2060

2080

2100 2120 Year

2140

2160

2180

2200

Fig. 4. Comparison of the cumulative collision probabilities of the benchmark and two ADR scenarios. Each curve is the average of 100 LEGEND Monte Carlo runs.

Table 1 The effective reduction factors (ERFs) for the 10 cm and larger LEO populations from the two ADR scenarios. PMD+ADR02 Number of LEO objects removed via ADR through 2206 (A) Reduction of LEO objects by 2206 (B) ERF by 2206 = (B)/(A)

PMD+ADR05

374

935

3899 10.4

7196 7.7

The PMD scenario is used as the benchmark for comparison. The “reduction of LEO objects by 2206” is the LEO population difference between the specified scenario and the benchmark for the year 2206.

titudes, where massive payloads and rocket bodies currently reside, and higher collision probabilities are expected. Even without specifying altitude, the removal criterion based on mass and collision probability effectively reduces the population growth in that critical altitude regime. With a removal rate of two objects per year, the projected LEO environment (dotted curve) is significantly lower than that predicted by the PMD scenario. With a removal rate of five objects per year, the LEO environment in 200 years, shown as the thin curve with diamonds, is similar to the current one. 4. Concluding remarks

the end of 2206 will be reduced by 10 objects (using the PMD prediction as a benchmark). This reflects a very high benefit ratio. It demonstrates that Eq. (1) is indeed an objective and effective selection criterion to identify objects which have the greatest potential of contributing to the growth of the future debris population. If objects that do not contribute to future collision activities are selected for removal, then the benefit ratio is no more than one. The ERF for the PMD+ADR05 scenario is lower than that for the PMD+ADR02 scenario, a result due to the fact that not all objects contribute equally to the growth of the future debris population. This expected trend means that, on average, the benefits of removing the two worst debris contributors is higher than that of removing the next three worst debris contributors. It is another indication that the selection criterion of Eq. (1) does identify objects for removal in the proper order. The spatial density distributions for objects 10 cm and larger are compared in Fig. 5. The long, dashed curve is the environment at the beginning of the future projection. Results from the PMD simulations (bold curve) indicate that the fastest debris growth region is between 800 and 1000 km al-

Any future debris environment simulations must rely on certain assumptions. Those adopted in the present study fall into two basic categories. The first category consists of assumptions considered to be typical by most modeling groups, such as the repeats of an 8-year launch cycle and an average solar activity cycle. The second category consists of assumptions which appear to be reasonable. These include the start time of ADR (the year 2020), the 90% success rate of the implementation of the commonly-adopted mitigation measures, and the immediate removal of objects identified for ADR. Each assumption has a different impact on the simulation's outcome [8]. It is a straightforward task to repeat similar numerical simulations with different parameters to encompass what is needed to stabilize the future LEO debris environment. In all likelihood, based on similar launch traffic, solar activity cycle, and mitigation measures, the results should be similar to what is shown above, i.e., it would require removing about 5 objects per year, as opposed to, say, 50 per year, to keep the future LEO environment stable. If more objects are removed, then the future LEO debris population could be lower than what is in the current environment.

J.-C. Liou et al. / Acta Astronautica 66 (2010) 648 -- 653

653

LEO Environment (objects >10 cm) 8.E-08 2206 PMD

7.E-08 Spatial Density (obj/km3)

2206 PMD + ADR02 2206 PMD + ADR05

6.E-08

2007 (Jan)

5.E-08 4.E-08 3.E-08 2.E-08 1.E-08 0.E+00 200

400

600

800

1000

1200

1400

1600

1800

2000

Altitude (km) Fig. 5. Comparison of the spatial density distributions for objects 10 cm and larger. The dashed curve is the environment at the beginning of the future projection.

As the international space community continues to work together to limit the generation of orbital debris, a more aggressive approach, active debris removal, must be seriously considered to remediate the environment. It is shown in this paper that simple and objective target selection criteria and removal strategies can be developed to make ADR an effective means to stabilize, or even reduce, the future LEO debris population. A combination of mitigation measures and active debris removal appears to be the only way to preserve the near-Earth environment for future generations. However, there are still many challenges ahead. Detailed cost analyses must be performed. Then, new ADR policies or guidelines will have to be developed in conjunction with the development of viable removal techniques. Finally, the source of funding needs to be identified, and the legal and liability issues must be addressed as well. References [1] Anon, Report of the Scientific and Technical Subcommittee of the United Nations Committee on the Peaceful Uses of Outer Space on the 44th session, held in Vienna from 12–23 February 2007, A/AC.105/890, (Annex IV), 6 March 2007.

[2] Anon, Endorsement by the General Assembly: Resolution adopted by the General Assembly, A/RES/62/217, 10 January 2008. [3] J.-C. Liou, N.L. Johnson, Risks in space from orbiting debris, Science 311 (2006) 340–341. [4] J.-C. Liou, N.L. Johnson, Instability of the present LEO satellite populations, Adv. Space Res. 41 (2008) 1046–1053. [5] D.J. Kessler, B.G. Cour-Palais, Collision frequency of artificial satellites: the creation of a debris belt, JGR 83 (A6) (1978) 2637–2646. [6] N.L. Johnson, E. Stansbery, J.-C. Liou, M. Horstman, C. Stokely, D. Whitlock, The characteristics and consequences of the break-up of the Fengyun-1C spacecraft, Acta Astronaut. 63 (2008) 128–135. [7] Anon, Four satellite breakups in February add to debris population, Orbital Debris Quarterly News 11 (2) (2007) 3. [8] J.-C. Liou, N.L. Johnson, A sensitivity study of the effectiveness of active debris removal in LEO, Acta Astronaut. 64 (2009) 236–243. [9] J.-C. Liou, D.T. Hall, P.H. Krisko, J.P. Opiela, LEGEND—a threedimensional LEO-to-GEO debris evolutionary model, Adv. Space Res. 34 (5) (2004) 981–986. [10] J.-C. Liou, Collision activities in the future orbital debris environment, Adv. Space Res. 38 (9) (2006) 2102–2106. [11] N.L. Johnson, P.H. Krisko, J.-C. Liou, P.D. Anz-Meador, NASA's new breakup model of EVOLVE 4.0, Adv. Space Res. 28 (9) (2001) 1377–1384. [12] J.-C. Liou, N.L. Johnson, Characterization of the cataloged Fengyun-1C fragments and their long-term effect on the LEO environment, Adv. Space Res. 43 (2009) 1407–1415.