Geotech Geol Eng

https://doi.org/10.1007/s10706-023-02561-y

TECHNICAL NOTE

Stability of Tailings Dam Constructed by both Upstream

and Centerline Methods
Yulong Chen · Zuoan Wei

Received: 4 August 2022 / Accepted: 9 July 2023

© The Author(s), under exclusive licence to Springer Nature Switzerland AG 2023

Abstract In China, there are more than 12,000 tail- performed to analyze the seepage and displacement
ings ponds, and the number is still growing. With fields, stability and storage capability of the tailings
the rapidly increasing number of tailings ponds, the dams constructed by upstream and mixed methods,
safety issue has drawn more and more attention by respectively. Through the simulated results, the mixed
governments and local communities. The design and method was demonstrated to be more beneficial to the
construction of each tailings pond must meet the spe- practical safety management of this tailings pond than
cific requirements of a particular mining project and the upstream method.
the changing conditions with time. This paper pre-
sents a case study of the Dahongshan Longdu tail- Keywords Tailings dam · Stability · Upstream
ings pond which combined upstream and centerline method · Centerline method · Mixed method
methods for dam construction. The Dahongshan Cop-
per Mine and Iron Mine were facing a challenge of
the disposal of increased production of tailings. The 1 Introduction
current upstream method of dam construction could
no longer guarantee its safety. Therefore, the origi- The Tailings dam is an important facility for the pro-
nal design of tailings pond should be revised for the duction of mining industry to store unwanted and cur-
safety of the tailings pond. This tailings dam used rently uneconomical mine waste tailings. However,
centerline method for subsequent construction after the mining industry are faced with unprecedented
total height of this dam was constructed up to 100 m challenges of deciding where to build a new tailings
by upstream method. This tailings pond initially used pond. The main reasons for experiencing stress are,
upstream method for the dam construction. To pile the tightening environmental protection legislation
up the increased production of tailings, the dam con- and regulations on tailings disposal, the rising cost
struction method was switched to centerline method of land acquisition, and the lessening suitable places
after dam height of 100 m. Numerical simulation was for the tailings pond construction near the mines.
To ensure the sustainable production of the mine
in case of being shut down due to no place to dis-
Y. Chen (*)
China Coal Research Institute, Beijing 100013, China charge tailings, more and more mine enterprises tend
e-mail: [email protected] to heighten the operating tailings dams to improve
their storage capacity (Wei et al. 2016). In this way,
Y. Chen · Z. Wei
the design height and the storage capacity of the tail-
School of Resources and Safety Engineering, Chongqing
University, Chongqing 400030, China ings pond will be increased. As a result, the risk of

Vol.: (0123456789)
13
 Geotech Geol Eng

tailings dam failure substantially increases accord- Longdu tailings pond to serve for the two mines.
ingly. Once dam failure occurs, heavy casualties, con- The tailing pond is located in a valley 500 m away
siderable property damage, and irreversible pollution from the East Bank of Jiasa River, Jiasa Town, Xin-
in downstream areas will be induced (Wei et al. 2009; ping County, Yuxi City, Yunnan Province, southwest-
Coimbra et al. 2019; de Lima et al. 2020; dos Santos ern China (Fig. 1). The distance between the tailings
Vergilio et al. 2021; Chen et al. 2023). Therefore, the pond and Dahongshan Copper Mine is about 8 km.
safety of tailings dams has become a widespread con- The tailings pond is valley-type and constructed by
cern and is under scrutiny by governments and local the upstream method. It has been put into service
communities (Ishihara et al. 2015; Chen et al. 2018; since July 1997. At first Dahongshan Copper Mine
Jin et al. 2018). started to discharge tailings to the tailings pond. After
Given that the upstream method, which is the main that Dahongshan Iron Mine began to pour tailings in
method in mine tailing’s disposal in China (Wei et al. 2003.
2009, 2012; Zhang et al. 2015), is inherently of lower Owing to the great demand for mineral resources,
dam stability and higher phreatic line than other the coal production of Dahongshan Copper Mine and
methods, it is difficult to meet the requirements of sta- Iron Mine was continuously increased, as well as the
bility for the high dam constructed by the upstream mineral processing quantity. The design capacity of
method. In contrast, centerline line method is more mill processing of Dahongshan Copper Mine was
suitable for high tailings dam construction because of increased from 1,584,000 t/a to 4,950,000 t/a, that of
lower phreatic line and higher stability of the dam. Dahongshan Iron Mine was increased from 4,500,000
Therefore, Longdu tailings dam, belonging to high t/a to 13,000,000 t/a, which largely exceeded the orig-
tailings dam, was designed to be further heightened inal design capacity of mill processing. Consequently,
to extend its tailings storage capability by centerline much more tailings were produced.
method in place of upstream method. That is, mixed In order to meet the needs of tailings storage, the
method of combining both upstream and centerline capacity of Longdu tailings pond was designed to
methods was used for dam construction. In particu- increase up to 120 million m ­ 3. The final height of
lar, first the upstream method was used, and then after tailings dam was designed to be 210 m at an eleva-
reaching a certain height the centerline line method tion of + 760 m. According to the storage capacity
was used. The mixed method of dam construction and dam height, the tailings pond was classified as
absorbs the advantages of both upstream and center- high dam. For storing such vast quantity of tailings,
line methods, and is beneficial to the stability of high the raising rate of the tailings dam was speeded up
tailings dams. to 10 m/a, which was much faster than the normal
This study presents a case study on heightening an raising rate. As the consolidation of tailings requires
operating tailings dam by mixed construction method time, the increment in dam raising rate leads to less
for the Longdu tailings pond in the Yunnan province time for the consolidation of tailings, thereby decreas-
of China. Both upstream and centerline methods were ing strength of the tailings.
combined for the dam construction. Comprehensive Moreover, as the underground backfilling require-
stability analyses of the tailings pond are performed. ment and mineral processing technology improved,
The seepage and displacement fields, stability and the discharged tailings have more fine particles less
storage capacity of the tailings pond constructed by than 0.03 mm and less coarse particles larger than
upstream method are compared with those by mixed 0.074 mm. It is noted that the fine tailings have much
method. The presented results manifest the advan- lower shear strength and lower permeability than the
tages of the mixed method to upstream method. coarse tailings. If upstream method is used for the
disposal of these fine tailings, the tailings dam may
have stability problems (Chandler and Tosatti, 1995;
2 Project Overview Vick, 1996), leading to environmental disasters (Ben-
zaazoua et al. 2008; Gidarakos et al. 2008).
In 1996, Dahongshan Iron Mine of Kunming Steel For these reasons, the stability of the Longdu tail-
Company and Dahongshan Copper Mine of Yuxi ings dam was considerably reduced, which at least
Mine Company jointly funded the construction of affects the sustainable operation of this pond, and

Vol:. (1234567890)
13
Geotech Geol Eng

Fig. 1  Location of the

Longdu tailings pond

even more would lead to dam failure accidents. Con- and comprehensive investigations, it should be the
struction by upstream method could no longer ensure best choice for the Longdu tailing pond to stop cur-
the stability of the tailings dam. Such dilemma forced rent upstream method to use centerline method for
the Dahongshan Copper Mine and Iron Mine to cre- subsequent dam construction to store increased out-
ate a new tailings pond. However, there was no suit- put of tailings. In other words, construction method
able area for a new tailings pond within 10 km around was switched from upstream method to centerline
the Dahongshan Mining area. To meet the storage method to heighten the tailings dam (Fig. 2). Ini-
capacity of tailings pond and to ensure its safety, it tially the upstream method was used for dam con-
is necessary to timely study the construction method struction. The starter dam was made up of perme-
of Longdu tailings dam. After careful considerations able rockfill, and had crest elevation of + 550 m, dam

Vol.: (0123456789)
13
 Geotech Geol Eng

Fig. 2  Longdu tailings dam

constructed by upstream
and centerline methods

height of 30 m, and upstream and downstream slopes of the tailings dam was 100 m. By the end of 2010,
of 1 V:2H. The dam wall slope was designed to be the elevation of tailings dam reached approximately
1 V:5H, and the final height was 70 m. Each sub-dam 620 m (Fig. 3).
was 2 m in height and constructed with coarse sands Since the beginning of 2011, the centerline
excavated from the tailings beach. The total height method has been used from the elevation of dam crest

Fig. 3  Longdu tailings pond by upstrea4 m method

Vol:. (1234567890)
13
Geotech Geol Eng

of + 620 m. Each sub-dam was designed to be 10 m

in height and 20 m in width, and had a downstream
slope of 1 V:4H. Finally, the tailings dam would be
built up to the design elevation of + 760 m. So, the
total height of the tailings dam would be 240 m.

3 Numerical Simulation

After investigating the Longdu tailings pond site,

a typical section profile 2–2’ was selected for this
numerical modelling, as shown in Fig. 4. In addi-
tion to solving the practical engineering problems
for Longdu tailings pond, this paper intends to com-
pare the seepage and displacement fields, stability,
and storage ability of the tailings dam constructed by
upstream and mixed methods. Therefore, the seepage
and displacement fields and stability of the tailings
dam by the upstream and mixed methods were mod-
elled respectively.
As shown in Fig. 5, the tailings pond was con-
Fig.  Layout of tailings dam and cross Sect. 2-2 for numerical structed by upstream method until dam height up
simulation

Fig. 5  Cross section of

tailings dam by upstream
method

Fig. 6  Cross section of tail-

ings dam by both upstream
and centerline methods

Vol.: (0123456789)
13
 Geotech Geol Eng

Fig. 7  Calculation model by upstream method (elevation + 620 m)

Fig.8  Calculation model by mixed method (elevation + 760 m)

Table 1  The properties of Material Dry unit Shear strength Coefficient of perme-
the tailings weight γ (KN/ ability (cm/s)
m 3)
Cohesion c (KPa) Internal fric- Horizontal Verti-
tion angle KH (× ­10−3) cal KV
ϕ (°) (× ­10−3)

① Bedrock / / / / /
② Starter dam 21.56 0.0 38.0 2.70 2.70
③ silty sand 20.50 15.00 32.00 0.45 0.80
④ Silty soil 21.00 22.00 30.00 0.40 0.50
⑤ Coarse tailings 20.97 15.00 34.80 0.45 0.80
⑥ Fine tailings 19.00 20.00 9.00 0.06 0.06
⑦ Silty clay 20.00 35.00 15.00 0.02 0.02
⑧ Rockfill 19.00 13.50 37.50 2.70 2.70

to 70 m at an elevation of + 620 m. From elevation and their properties were measured by laboratory
of + 620 m to + 760 m, this pond was constructed testing, and the results are summarized in Table 1.
by centerline method, as shown in Fig. 6. The final
height would be 210 m.
The seepage and displacement fields were ana-
lyzed by ANSYS. For this stability analysis, Slide 4 Results
(v 5.0) software was used. The calculation mod-
els were built as shown in Fig. 7 for the pond by 4.1 Seepage Field
upstream method and Fig. 8 for that by mixed
method. The dry beach length was considered to be Figures 9 and 10 show the simulated seepage field
150 m for the both ponds by upstream and mixed of Longdu tailings dam by upstream and mixed
methods. The tailings were sampled from the field methods, respectively. The depth of the phreatic
line of the pond by mixed method was obviously

Vol:. (1234567890)
13
Geotech Geol Eng

Fig. 9  Seepage field of Longdu tailings pond by upstream method

lower than that by upstream method. This was more 4.2 Displacement Field
conducive to the stability of the dam. Therefore,
the simulated results of the seepage field indicated The simulated displacement field is shown in Fig. 11
that the mixed method was better than the upstream for the pond by upstream method and Fig. 12 for that
method. by mixed method. Similar displacement distributions
were observed in both tailings pond by upstream
and mixed methods. Both Figs. 11, 12 show that the

Vol.: (0123456789)
13
 Geotech Geol Eng

Fig. 10  Seepage field of Longdu tailings pond by mixed method

horizontal displacement was small, and it was toward 4.3 Stability

the downstream of the dam. The vertical displace-
ment was large, and the highest value appeared at the Stability analysis of the tailings dam was carried
dam crest, indicating that the compressibility of the out by using ordinary slices method and Bishop’s
tailings was rather large so that the settlement was simplified method, in accordance with the national
quite significant. technical code. The stability of the tailings dam

Vol:. (1234567890)
13
Geotech Geol Eng

Fig. 11  Displacement field

of Longdu tailings pond by
upstream method

Fig. 12  Displacement field

of Longdu tailings pond by
mixed method

constructed up to elevation of + 620 m by upstream constructed by upstream method until the elevation
method was calculated. In order to compare with the of + 760 m. Two sets of stability analysis results are
results of tailings dam by mixed method, stability tabulated in Table 2.
analysis was carried out for the dam continuously

Vol.: (0123456789)
13
 Geotech Geol Eng

Table 2  Factors of safety Factor of safety Upstream method Mixed method

calculated for tailings dam
constructed by upstream Ordinary slices Bishop’s simpli- Ordinary slices Bishop’s
and mixed methods method fied method method simplified
method

Cross Sect. 2-2 at + 620 m 1.740 1.787 / /

Cross Sect. 2-2 at + 760 m 1.277 1.422 1.582 1.625
Criteria required by Chinese 1.25 1.35 1.25 1.35
Standards

Fig. 13  Stability analysis of the tailings dam by mixed method

It can be seen that, for the elevation of + 620 m, national code. Besides, the factors of safety of the
the factors of safety were far higher than that values dam constructed by mixed method were higher than
required by the national code. When the dam eleva- that by upstream method at the same dam height.
tion increased to + 760 m, the factors of safety were Therefore, the mixed method was more conducive to
greatly decreased, which were just marginally higher the dam stability than the upstream method.
than those values required by the national code. It The centerline method has been used to construct
indicated that the upstream method could satisfy the the tailings pond from elevation of + 620 m since
stability required by the national code. 2011. Figure 14 shows the scenario in course of
When the tailings pond was subsequently con- construction. Site observations have shown that the
structed up to elevation of + 760 m by centerline mixed method has been successful in dealing with
method, its stability was then calculated. The sliding tailings dam safe heightening.
surface was determined as shown in Fig. 13 and fac-
tors of safety were listed in Table 2. The factors of 4.4 Storage Capacity
safety for the tailings pond by mixed method were
much higher than that values required by the national Figure 15 illustrates the cross section of the tail-
code, indicating that its stability could satisfy the ings pond combining upstream and centerline

Vol:. (1234567890)
13
Geotech Geol Eng

Fig. 14  Longdu tailings pond being constructed by centerline method

Fig. 15  Storage capacity of

the tailings pond compar-
ing upstream method with
mixed method

methods. The red zone of quadrilateral OABC refers Therefore, the original design of tailings pond should
to the incremental storage capacity of tailings pond be revised for the safety of the tailings pond. This
by mixed method to that by upstream method. The tailings dam used centerline method for subsequent
increased storage capacity of tailings pond by mixed construction after total height of this dam was con-
method is 3.3 × ­104 ­m3/m. When dam axis is 200 m structed up to 100 m by upstream method.
long, the storage capacity of tailings pond by mixed Based on the numerical simulation results, sta-
method would increase by 6.6 × ­106 ­m3. bility analyses of this tailings dam by upstream and
mixed methods were obtained. The seepage and dis-
placement fields, stability and storage capacity of
5 Conclusions the tailings dam constructed by upstream method
were compared with those by mixed method. The
The Dahongshan Copper Mine and Iron Mine were presented results showed that the mixed method was
facing a challenge of the disposal of increased pro- more beneficial to the practical safety management of
duction of tailings. The current upstream method of this tailings pond than the upstream method.
dam construction could no longer guarantee its safety.

Vol.: (0123456789)
13
 Geotech Geol Eng

Acknowledgements This study was financially supported the Córrego do Feijão tailings dam disaster, Brumadinho,
by National Natural Science Foundation of China (52009131, Brazil. Landslides 17:2895–2906
52004285). dos Santos VC, Lacerda D, da Silva ST et al (2021) Immediate
and long-term impacts of one of the worst mining tailing
Author contributions YC performed the numerical simula- dam failure worldwide (Bento Rodrigues, Minas Gerais,
tion and wrote this manuscript. ZW supervised this study. Brazil). Sci Total Environ 756:143697. https://​doi.​org/​10.​
1016/j.​scito​tenv.​2020.​143697
Funding Funding was provided by National Natural Science Gidarakos E, Anastasiadou K, Koumantakis E, Nikolaos S
Foundation of China, (52009131), Yulong Chen (2008) Investigative studies for the use of an inactive
asbestos mine as a disposal site for asbestos wastes. J
Data availability The datasets used and analyzed during the Hazard Mater 153:955–965. https://​doi.​org/​10.​1016/j.​
current study are available from the corresponding author on jhazm​at.​2007.​09.​060
reasonable request. Ishihara K, Ueno K, Yamada S et al (2015) Breach of a tail-
ings dam in the 2011 earthquake in Japan. Soil Dyn
Declarations Earthq Eng 68:3–22. https://​doi.​org/​10.​1016/j.​soild​yn.​
2014.​10.​010
Conflict of interest The authors declare no competing inter- Jin J, Song C, Liang B et al (2018) Dynamic characteristics
est. of tailings reservoir under seismic load. Environ Earth
Sci 77:1–11. https://​doi.​org/​10.​1007/​s12665-​018-​7836-1
Consent for publication Not applicable. Wei Z, Yin G, Li G et al (2009) Reinforced terraced fields
method for fine tailings disposal. Miner Eng 22:1053–
Ethics approval and consent to participate Not applicable. 1059. https://​doi.​org/​10.​1016/j.​mineng.​2009.​03.​014
Wei Z, Yin G, Wang JG et al (2012) Stability analysis and
supporting system design of a high-steep cut soil slope
on an ancient landslide during highway construction
References of Tehran-Chalus. Environ Earth Sci 67:1651–1662.
https://​doi.​org/​10.​1007/​s12665-​012-​1606-2
Wei Z, Yin G, Wan L, Li G (2016) A case study on a geotech-
Benzaazoua M, Bussière B, Demers I et al (2008) Integrated nical investigation of drainage methods for heightening a
mine tailings management by combining environmental tailings dam. Environ Earth Sci 75:1–10. https://​doi.​org/​
desulphurization and cemented paste backfill: application 10.​1007/​s12665-​015-​5029-8
to mine Doyon, Quebec, Canada. Miner Eng 21:330–340. Zhang Q, Yin G, Wei Z et al (2015) An experimental study of
https://​doi.​org/​10.​1016/j.​mineng.​2007.​11.​012 the mechanical features of layered structures in dam tail-
Chen Y, Wei Z, Irfan M et al (2018) Laboratory investiga- ings from macroscopic and microscopic points of view.
tion of the relationship between electrical resistiv- Eng Geol 195:142–154. https://​doi.​org/​10.​1016/j.​enggeo.​
ity and geotechnical properties of phosphate tailings. 2015.​05.​031
Measurement 126:289–298. https://​doi.​org/​10.​1016/j.​
measu​rement.​2018.​05.​095
Chen Y, Jing X, Wei Z, Wang M (2023) Physical and numeri- Publisher’s Note Springer Nature remains neutral with regard
cal modeling of the hypothetical tailings dam breach to jurisdictional claims in published maps and institutional
runout and mitigation with a slurry-resisting barrier. affiliations.
Bull Eng Geol Environ 82:1–19. https://​doi.​org/​10.​1007/​
s10064-​023-​03312-2 Springer Nature or its licensor (e.g. a society or other partner)
Coimbra KTO, Alcântara E, de Souza Filho CR (2019) An holds exclusive rights to this article under a publishing
assessment of natural and manmade hazard effects on the agreement with the author(s) or other rightsholder(s); author
underwater light field of the Doce River continental shelf. self-archiving of the accepted manuscript version of this article
Sci Total Environ 685:1087–1096. https://​doi.​org/​10.​ is solely governed by the terms of such publishing agreement
1016/j.​scito​tenv.​2019.​06.​127 and applicable law.
de Lima RE, de Lima PJ, da Silva AF, Acordes FA (2020) An
anthropogenic flow type gravitational mass movement:

Vol:. (1234567890)
13