Influence of MgO and CaCl₂ on the rheological properties of bentonitic clays from the new Paraíba-Brazil deposits using experimental planning and statistical analysis

Influência do MgO e CaCl ₂ nas propriedades reológicas de argilas bentoníticas dos novos depósitos da Paraíba-Brasil utilizando planejamento experimental e análise estatística

Abstract

New deposits of bentonite clays have been discovered in the Brazilian State of Paraíba; the most recent was at the municipality of Olivedos. Recent studies have discovered the presence of high levels of non-clay minerals that can produce unsatisfactory results when attempting to use these clays in drilling fluids. In order to make them suitable for this purpose, the MgO and CaCl₂ as chemical additives were used and their influences on the rheological properties of these clays were analyzed, using an experimental planning technique and statistical analysis. The samples were obtained using experimental modeling by the delineation of mixtures technique; first, the clays were transformed with sodium carbonate and then dosed with MgO and CaCl₂. The rheological properties, apparent viscosity (AV) and plastic viscosity (PV) were determined according to the Petrobras standard (AV≥15 cP; PV≥4 cP). The results showed that the values of AV and PV increased considerably and that MgO was the additive that contributed most to the improvement of these properties, making these additives suitable for use in water-based drilling fluids.

Keywords:
Olivedos’ bentonites; additives; drilling fluids

Resumo

Novos depósitos de argila bentonítica foram descobertos no estado brasileiro da Paraíba, sendo o mais recente no município de Olivedos. Estudos recentes descobriram a presença de altos níveis de minerais não argilosos que podem produzir resultados insatisfatórios ao tentar usar essas argilas em fluidos de perfuração. Para torná-los adequados para esse fim, utilizaram-se MgO e CaCl₂ como aditivos químicos e analisaram-se suas influências nas propriedades reológicas dessas argilas, utilizando técnica de planejamento experimental e análise estatística. As amostras foram obtidas utilizando modelagem experimental pela técnica de delineamento de misturas; inicialmente as argilas foram transformadas com compostos de sódio e posteriormente dosadas com MgO e CaCl₂. As propriedades reológicas, viscosidade aparente (AV) e viscosidade plástica (PV), foram determinadas de acordo com a normativa da Petrobras (AV≥15 cP; PV≥4 cP). Os resultados mostraram que os valores de AV e PV aumentaram consideravelmente e que o MgO foi o aditivo que mais contribuiu para a melhoria dessas propriedades, tornando-os adequados para uso em fluidos de perfuração à base de água.

Palavras-chave:
bentonitas de Olivedos; aditivos; fluidos de perfuração

INTRODUCTION

Paraíba is the Brazilian State containing several bentonite clay deposits, mainly in the Boa Vista municipality. These deposits have already reached the exhaustion phase. Recent studies ^1)-(4 have been developed using bentonites from newly discovered deposits at the Olivedos, Cubati, Pedra Lavrada, and Sossego municipalities, but the presence of non-clay minerals such as mica and quartz render them unsuitable for use in drilling fluids. Because of this difficulty, several companies in Paraíba have begun to import bentonitic clays from countries such as Turkey and India, in order to preserve the continuity of their production processes. A viable alternative to using these Paraíba’s clays is to treat them with sodium carbonate (Na₂CO₃) isolated or combined with secondary chemical additives to find intermediate gels between the flocculation and deflocculation states that characterize water-based drilling fluids.

According to van Olphen ⁵, drilling fluids are clay-water systems having an intermediate behavior between flocculated and deflocculated state, which are obtained naturally from Wyoming sodium clays that provide the apparent and plastic viscosities originally used by API (American Petroleum Institute) as a worldwide standard. The deflocculated systems have high apparent and plastic viscosities (AV and PV), whereas the flocculated systems can be divided into two types: 1) flocculated gel (as termed in ⁵) with high apparent viscosity and low plastic viscosity, having anisometric clay particles forming a ‘face to edge’ house of cards micro-configuration; and 2) flocculated with phase separation, where the anisometric clay particles have a ‘face to face’ micro-configuration. In these extreme cases, secondary additives (also called chemical additives) can be used to transform this clay-water system into a true drilling fluid suitable for the specific geological conditions found at each drilling site, and is a vast and complex subject covered by thousands of patents and industrial secrets.

Several studies ^6)-(11 have demonstrated that the use of ions, such as Ca²⁺, Mg²⁺, K⁺, or Li²⁺ as secondary chemical additives in clay-water systems, can transform them in systems ranging from flocculated to deflocculated, according to the proportion and combination of ions added. These systems have been shown to have satisfactory rheological property values at intermediate points, where traditional techniques do not indicate adequate results. Some researchers ^12)-(14 have used statistical and mathematical techniques to study the effect of clay compositions and additives on the rheological properties of clay-water systems. This methodology has found application in a number of technological areas ^13)-(17. In the cases mentioned, this methodology led to better results and optimized the systems with a minimum of experiments. However, there is a lack of knowledge regarding the use of these systems with clays from the newly discovered deposits mentioned above. In order to improve the rheological behavior of the new bentonite deposits in the Paraíba State, Brazil, that contain a high amount of non-clay minerals leading to unsatisfactory results using traditional tests and the Petrobras standard ¹⁸, in this study CaCl₂ and MgO flocculants were investigated as secondary chemical additives to make possible the use of the new Olivedos-Paraíba clays in water-based drilling fluids using the mixture delineation technique and statistical analysis.

MATERIALS AND METHODS

The clays studied, named AM1 and AM2, were obtained from samples of the new deposits in the Olivedos municipality, Paraíba-Brazil, and the average samples were obtained using the delineation of mixture experiment modeling technique ¹³. Among the samples studied in ¹⁹, those from the Olivedos municipality showed the most promising results when mixed with Chocolate clay from the Boa Vista municipality. In this study, Olivedos clays were also investigated individually, which did not produce satisfactory results, driving the decision to use secondary additives with these samples. The physical, chemical and mineralogical characterizations of the used clay samples are described in ¹², which shows that they had clay minerals from the smectite group with kaolinite and mica, and quartz as a non-clay mineral. These clays were transformed with a sodium compound through the classic process with the addition of sodium carbonate at a ratio of 125 meq/100 g of dry clay. The results of the rheological tests, according to the Petrobras standard ¹⁸, were unsatisfactory. The following compounds were used as additives: sodium carbonate (Na₂CO₃, Vetec); 99.0% pure calcium chloride (CaCl₂); and 99.0% pure magnesium oxide (MgO), both supplied by Casa da Química.

The clays were dried in an oven at approximately 60 °C, fragmented in a ball mill and sieved with an ABNT No. 200 (0.074 mm) sieve. After processing, the polycationic natural clays were transformed with sodium carbonate as described above and then cured for a period of 5 days to allow a complete exchange of cations. Following this, MgO and CaCl₂ were added using the delineation of the mixture experiment modeling technique ²⁰. The dispersions were prepared and AV and PV parameters were determined according to Petrobras standard ¹⁸. To define the compositions, simplex centroid network planning {3,2} was used, with interior points added, totaling ten experimental runs. These experiments were performed randomly and each test was conducted in duplicate. The proportions of the components ranged from 97.5% to 100% for clay and from 0% to 2.5% for MgO and CaCl₂. The compositions are shown in Table I. The results were obtained in duplicate and the statistical analysis was performed with Statistica v.6.0 software.

Table I
Proportions of components (%) of simplex centroid network planning {3,2} with interior points added.
Tabela I
Proporções dos componentes (%) do planejamento em rede simplex centroide {3,2} aumentado com pontos interiores.

Composition	Clay (AM1 or AM2)	MgO	CaCl2
1	100.00	0.00	0.00
2	97.50	2.50	0.00
3	97.50	0.00	2.50
4	98.75	1.25	0.00
5	98.75	0.00	1.25
6	97.50	1.25	1.25
7	99.16	0.42	0.42
8	97.91	1.67	0.42
9	97.91	0.42	1.67
10	98.34	0.83	0.83

RESULTS AND DISCUSSION

Table II shows the results of rheological parameters obtained for the dispersions prepared with the compositions listed in Table I. From the results obtained, including duplicates, regression equations were generated (Table III), correlating the mass proportion of clay and MgO and CaCl₂ additives with the apparent viscosity (AV) and plastic viscosity (PV). The equations presented were statistically significant at the 95% confidence level and described the behavior of the dispersions’ properties as a function of the component proportions. Table IV presents the main statistical parameters obtained from the analysis of variance of the models presented in Table III. The statistical parameters indicated that the models were well-adjusted. The p-values showed that the models were statistically significant at the stipulated level (p-value level of significance). The relationships between the calculated and tabulated F-test values for AV₁, PV₁, and AV₂ showed that the models were not only significant but also predictive. For PV₂, it can be stated that the model was significant ^{14), (21}.

Table II
Rheological parameters of the compositions studied.
Tabela II
Parâmetros reológicos das composições estudadas.

Composition	AV1 (cP)	PV1 (cP)	AV2 (cP)	PV2 (cP)
1	19.75	1.00	13.50	3.00
	17.00	2.00	12.00	3.00
2	17.50	7.00	27.50	6.00
	20.00	6.00	26.00	7.00
3	11.00	4.00	17.00	4.00
	11.50	4.00	17.00	5.00
4	19.00	4.00	24.50	5.00
	18.00	4.00	22.50	4.00
5	11.00	3.00	14.00	3.00
	10.00	4.00	13.00	3.00
6	17.50	6.00	31.50	9.00
	18.00	5.00	29.50	8.00
7	20.00	4.00	22.00	4.00
	20.00	3.00	24.00	4.00
8	17.50	3.00	20.00	5.00
	20.50	5.00	22.00	4.00
9	13.00	5.00	18.00	4.00
	15.50	4.00	18.50	4.00
10	20.00	5.00	26.50	6.00
	20.00	6.00	24.50	5.00
Specification EP-1EP-00011-A (19) API 13A: AV≥15cP; PV≥4cP

• AV₁ and PV₁ - AV (apparent viscosity) and PV (plastic viscosity) for compositions formulated with sample AM₁; AV₂ and PV₂ - AV and PV for compositions formulated with sample AM₂.

Table III
Decoded regression equations for AV₁, PV₁, AV₂, and PV₂.
Tabela III
Equações de regressão decodificadas para AV₁, PV₁, AV₂ e PV₂.

Variable	Equation
AV1	AV1=0.19.AM1-0.04.B+138.91.C-1.46.AM1.C+0.03.AM1.B.C
PV1	PV1=0.02.AM1+1.69.B+1.02.C
AV2	AV2=0.1.AM2+134160.3.B+1.6.C-2039.0.AM2.B-675.1.B.C+7.0.AM2.B.(AM2-B)
PV2	PV2=0.03.AM2+1.33.B+0.53.C+1.54.B.C

• AM₁, AM₂, B, and C represent clay AM₁, clay AM₂, MgO, and CaCl₂, respectively.

Table IV
Statistical parameters of the analyses of variance for the AV and PV variables relative to the selected models.
Tabela IV
Parâmetros estatísticos das análises de variância das variáveis viscosidade aparente (AV) e viscosidade plástica (PV) relativos aos modelos escolhidos.

Variable	Model	Test Fcal	Fcal/Ftab	p-value	R2
AV1	Special cubic	21.0972	6.8945	0.0000	0.8491
PV1	Linear	16.7879	4.6762	0.0001	0.6639
AV2	Complete cubic	35.3216	11.9330	0.0000	0.9266
PV2	Quadratic	12.8902	3.9784	0.0002	0.7073

• F_cal - calculated F-test (Fisher test); F_cal/F_tab - ratio between the calculated F-test and the tabulated F-test; p-value - descriptive level or probability of significance; R² - correlation coefficient.

Figs. 1 and 2 illustrate the response surfaces and contour plots for AV₁ and PV₁ values, respectively. It was observed that MgO, as Mg(OH)₂, was the component that most contributed to the increase of AV, probably because the Mg²⁺ cation acted as a bridging bond that brought the clay particles closer together, leading to a greater tendency to flocculate. This ion, even when hydrated, decreases the thickness of the adsorbed water layer due to its higher charge and smaller ionic radius ^{7), (14} that approximated the clay particles

Figure 1:
Response surface (a) and contour plot (b) for AV₁ calculated from the special cubic model as a function of the quantity of AM₁ clay, MgO, and CaCl₂.
Figura 1:
Superfície de resposta (a) e curvas de nível (b) para AV₁ calculadas a partir do modelo cúbico especial em função da quantidade da argila AM₁, MgO e CaCl₂.

Figure 2:
Response surface (a) and contour plot (b) for PV₁ calculated from the linear model as a function of the quantity of AM₁ clay, MgO, and CaCl₂.
Figura 2:
Superfície de resposta (a) e curvas de nível (b) para PV₁ calculadas a partir do modelo linear em função da quantidade da argila AM₁, MgO e CaCl₂.

According to Sousa Santos ²¹, the addition of bivalent cations decreases the electrokinetic potential of the clay particles and, therefore, the repulsion between these particles. As a consequence, particle agglomerates are formed, that is, the phenomenon of flocculation occurs. Polyvalent cations replace sodium on the surface of the clay particles, reducing and reversing their surface charge. After this reduction and that of the electrokinetic potential, the particles naturally associate ‘face to face’, producing the flocculated state with phase separation. The response surfaces for PV make it possible to conclude that, although both additives contributed to the increase of the plastic viscosity, MgO as Mg(OH)₂ was the component that more effectively increased this parameter. MgO in the Mg(OH)₂ form induced face-to-face interactions, resulting in the flocculated state, which favored both apparent viscosity (AV) and plastic viscosity (PV). In the flocculated gel state, it has high AV and low PV ^{7), (14}.

Fig. 3 shows the overlapping between the contour plots of parameters AV₁ and PV₁. The cross-hatched area indicated the compositions whose rheological parameters AV₁ and PV₁ met the requirements established by the Petrobras standard ¹⁸ for use as drilling fluids in water-based oil wells ¹⁴. In order to validate the model experimentally, compositions within the limits were selected. Using the equations from Table III, the predicted values for AV₁ and PV₁ were calculated and then determined experimentally ¹⁴. The results are shown in Table V. The predicted and experimentally-measured results for the AV₁ and PV₁ values were similar, making it possible to state that the statistical models were reliable for predictive purposes. Similar results were obtained in ¹³.

Figure 3:
Intersection between the AV₁ and PV₁ contour plots showing a cross-sectional area of compositions with AV₁15.0 cP and PV₁ 4.0 cP.
Figura 3:
Intersecção das curvas de níveis de AV₁ e PV₁ apresentando área hachurada de composições com AV₁15,0 cP e PV₁4,0 cP.

Table V
Predicted and experimentally measured values for rheological parameters AV₁ and PV₁.
Tabela V
Valores preditos e medidos experimentalmente para os parâmetros reológicos AV₁ e PV₁.

Composition	Component proportion (%)			Predicted value		Experimental value
	AM1	MgO	CaCl2	AV1 (cP)	PV1 (cP)	AV1 (cP)	PV1 (cP)
i	97.70	2.10	0.20	19.18	5.73	19.12	6.35
ii	97.74	0.79	1.47	16.40	4.81	17.85	4.25
iii	99.52	0.28	0.20	18.16	2.69	18.83	3.79
iv	98.30	0.95	0.75	20.39	4.36	19.03	4.58

Figs. 4 and 5 illustrate the response surfaces and contour plots for AV₂ and PV₂, respectively. The addition of MgO and CaCl₂ improved AV and PV over most of the composition region and contributed to their increase. Fig. 6 shows the overlap between the contour plots of parameters AV₂ and PV₂. The cross-hatched area indicated the compositions whose rheological parameters met the requirements established by the Petrobras standard ¹⁸ for use in water-based oil well drilling fluids. In order to experimentally validate the model, compositions within the determined limits were selected. Using the equations from Table III, the predicted values for AV₂ and PV₂ were calculated and then determined experimentally. The results are shown in Table VI. The predicted and experimentally-measured results for the values of AV₂ and PV₂ make it possible to conclude that the statistical models were reliable for predictive purposes. Similar results were obtained in ¹³.

Figure 4:
Response surface (a) and contour plot (b) for AV₂ calculated from the complete cubic model as a function of the amount of clay AM₂, MgO, and CaCl₂.
Figura 4:
Superfície de resposta (a) e curvas de nível (b) para AV₂ calculadas a partir do modelo cúbico completo em função da quantidade da argila AM₂, MgO e CaCl₂.

Figure 5:
Response surface (a) and contour plot (b) for PV₂ calculated from the quadratic model as a function of the amount of clay AM₂, MgO, and CaCl₂.
Figura 5:
Superfície de resposta (a) e curvas de nível (b) para PV₂ calculadas a partir do modelo quadrático em função da quantidade da argila AM₂, MgO e CaCl₂.

Figure 6:
Intersection between the AV₂ and PV₂ contour plots showing a cross-sectional area of compositions with AV₂15.0 cP and PV₂4.0 cP.
Figura 6:
Intersecção das curvas de níveis de AV₂ e PV₂ apresentando área hachurada de composições com VA₂15,0 cP e VP₂4,0 cP.

Table VI
Compositions used for the model tests and the predicted and experimentally measured values for rheological parameters AV₂ and PV₂.
Tabela VI
Composições utilizadas nos testes dos modelos e os respectivos valores observados e previstos de AV₂ e PV₂.

Composition	Component proportion (%)			Predicted value		Experimental value
	AM2	MgO	CaCl2	AV2 (cP)	PV2 (cP)	AV2 (cP)	PV2 (cP)
i	98.73	0.95	0.32	24.92	4.57	25.50	5.00
ii	98.14	.95	0.91	23.79	5.74	23.50	5.00
iii	97.74	1.94	0.32	22.88	6.35	22.50	6.00
iv	97.70	0.36	1.94	20.84	5.22	21.50	5.50

CONCLUSIONS

It was possible to conclude that the use of secondary additives MgO and CaCl₂ increased the apparent and plastic viscosity values of dispersions prepared with smectite clays from the Olivedos municipality, Paraíba State, Brazil. The results also showed that after the additivation process the clays studied became suitable for use in water-based drilling fluids and that the experimental planning technique proved to be an important and fundamental tool in the study of the optimization and modeling of clay compositions, avoiding the need to import raw materials for the extraction of petroleum in Brazil.

REFERENCES

• ¹
  R.R. Menezes, L.F.A. Campos, H.S. Ferreira, L.N. Marques, G.A. Neves, H.C. Ferreira, Cerâmica 55, 336 (2009) 349.
• ²
  I.D.S. Pereira, V.N.F. Lisboa, I.A. Silva, J.M.R. Figueiredo, G.A. Neves, R.R. Menezes, Mater. Sci. Forum 820 (2015) 65.
• ³
  I.A. Silva, I.D.S. Pereira, W.S. Cavalcante, F.K.A. Sousa, G.A. Neves, H.C. Ferreira, Mater. Sci. Forum 820 (2015) 68.
• ⁴
  B.M.A.B. Buriti, J.S. Buriti, I.A. Silva, J.M. Cartaxo, G.A. Neves, H.C. Ferreira, Cerâmica 65, 373 (2019) 78.
• ⁵
  H. van Olphen, Clay colloid chemistry: for clay technologists, geologists and soil scientists, John Wiley Sons, USA (1997).
• ⁶
  J.F. Duarte-Neto, J.M. Cartaxo, G.A. Neves, R.R. Menezes, Rev. Eletr. Mater. Proc. 9 (2014) 51.
• ⁷
  B.M.A. Brito, J.M. Cartaxo, N.F.C. Nascimento, H.C. Ferreira, G.A. Neves, R.R. Menezes, Cerâmica 62, 361 (2016) 45.
• ⁸
  I.A. da Silva, F.K.A. de Sousa, R.R. Menezes, H.S. Ferreira, G. de A. Neves, H.C. Ferreira, Cerâmica 64, 369 (2018) 109.
• ⁹
  A.E. Bayat, P.J. Moghanloo, A. Piroozian, R. Rafati, Colloids Surf. A 555 (2018) 256.
• ¹⁰
  R. Rafati, S.R. Smith, A.S. Haddad, R. Novara, H. Hamidi, J. Pet. Sci. Eng. 161 (2018) 61.
• ¹¹
  R.C. Balaban, E.L.F. Vidal, M.R. Borges, Appl. Clay Sci. 105-106 (2015) 124.
• ¹²
  B.M.A. Brito, P.M. Bastos, A.J.A Gama, J.M. Cartaxo, G.A. Neves, H.C. Ferreira, Cerâmica 64, 370 (2018) 254.
• ¹³
  P.M. Bastos, B.M.A. Brito, A.J.A. Gama, J.M. Cartaxo, G.A. Neves, L.F.A. Campos, Cerâmica 63, 366 (2017) 187.
• ¹⁴
  L.V. Amorim., M.I.R. Barbosa, H.L. Lira, H.C. Ferreira, Mater. Res. 10 (2007) 53.
• ¹⁵
  J. Li, R. Pan, B. Guo, J. Shan, J. Nat. Gas Sci. Eng. 17 (2014) 131.
• ¹⁶
  A.J.A. Gama, J.M.R. Figuerêdo, J.M. Cartaxo, M.A. Gama, G.A. Neves, H.C. Ferreira, Cerâmica 63, 367 (2017) 336.
• ¹⁷
  W. Zheng, H. Cao, J. Zhong, S. Qian, Z. Peng, C. Shen, J. Non-Cryst. Solids 409 (2015) 27.
• ¹⁸
  EP-1EP-00011-A, “Ensaio de viscosificante para fluidos base água na exploração e produção de petróleo”, Petrobras (2011).
• ¹⁹
  P.M. Bastos, “Otimização e modelagem de propriedades reológicas de argilas esmectitas do estado da Paraíba para uso em fluidos de perfuração de poços de petróleo”, Tese Dr., Un. Fed. Campina Grande (2017).
• ²⁰
  M.I. Rodrigues, A.F. Iemma, Planejamento de experimentos e otimização de processos: uma estratégia sequencial de planejamentos, Casa Pão Ed. (2005).
• ²¹
  P.S. Souza, Ciência e tecnologia de argilas, Edgard Blucher, S. Paulo (1989).

Publication Dates

• Publication in this collection
  13 Dec 2019
• Date of issue
  Jan-Mar 2020

History

• Received
  24 May 2019
• Reviewed
  29 July 2019
• Reviewed
  13 Sept 2019
• Accepted
  16 Sept 2019